[top]
NAME
yn_case -- munit test case for yn.
USAGE
run_tcase('yntor_tcase')
INPUTS
(none)
OUTPUTS
* tc -- tcase case struct (tcase_s)
AUTHOR
Charlie Cornish
CREATION DATE
2005-Feb-16
COPYRIGHT
(c) Charles R. Cornish 2005
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
yn -- Determine Boolean value. SYNOPSIS [tf] = yn(x, [return_format])
INPUTS
* x -- value to check (Boolean numeric or character string) * return_format -- (optional) format of returned value, see DESCRIPTION for details (string).
OUTPUTS
* tf -- Boolean result (Boolean numeric or character string).
SIDE EFFECTS
DESCRIPTION
yn evaluates the value of x and determines whether is a reduces to a Boolean (true/false) value. It is a handy utlity for converting between formats for Boolean values, i.e. T/F -> true/false -> 1/0 -> Y/N -> yes/no, as well as for evaluating expressions within the callers workspace. When x is numeric, the function checks the values of the array and returns a Boolean value depending on whether all values are zero (false) or otherwise (true). When x is a character string and has the values: * 'T', 'TRUE', 'YES', 'Y' (case-insensitive), it returns a value of 'true' * 'F', 'FALSE', 'NO', 'N' (case-insensitive), it returns a value of 'false' For other characters strings, the function attempts to evaluate the expression using the specified variables contained in caller's workspace. If the expression can not be evaluated, an error is thrown. If x is neither numeric or character, an error is thrown. The 'return_format' input argument controls format of the output argument. Valid values for 'return_format' are case-insenstive and include: * 'numeric' -- returns 1/0 for Boolean true/false * 'tf' -- returns T/F for Boolean true/false * 'truefalse' -- returns TRUE/FALSE for Boolean true/false * 'yn' -- returns Y/N for Boolean true/false * 'yesno' -- returns YES/NO for Boolean true/false. The default value of 'return_format' is 'numeric'.
USAGE
tf = yn(x) tf = yn(x, return_type)
ERRORS
WMTSA:InvalidNumArguments, WMTSA:InvalidStringArgumentValue, WMTSA:InvalidArgumentType, WMTSA:InvalidArgumentValue
EXAMPLE
tf = yn(1) % Returns tf = 1 tf = yn(1, 'tf') % Returns tf = 'T" tf = yn('F') % Returns tf = 0 tf = yn('T', 'yesno') % Returns tf = 'YES' x = 1; y = 0; tf = yn('x == y', 'truefalse') % Returns tf = 'FALSE' TOOLBOX wmtsa/utils CATEGORY WMTSA Utilities
AUTHOR
Charlie Cornish
CREATION DATE
2005-02-16
COPYRIGHT
(c) 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wvar_var_fd_sdf_acvs -- Calculate variance of wavelet variance for a FD process for given wavelet transform filter. SYNOPSIS
INPUTS
OUTPUTS
SIDE EFFECTS
DESCRIPTION
USAGE
WARNINGS
ERRORS
EXAMPLE
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wtfilter_tsuite -- munit test suite to test wtfilter and associated functions.
USAGE
run_tsuite('wtfilter_tsuite')
INPUTS
OUTPUTS
ts = tsuite structure for wtfilter transform testsuite.
DESCRIPTION
AUTHOR
Charlie Cornish
CREATION DATE
2005-03-02
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wtfilter_tcase -- munit test case to test wtfilter.
USAGE
run_tcase('wtfilter_tcase')
INPUTS
(none)
OUTPUTS
tc = tcase structure for wtfilter testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
wtfilter
AUTHOR
Charlie Cornish
CREATION DATE
2005-03-01
COPYRIGHT
(c) Charles R. Cornish 2005
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wtfilter -- Define wavelet transform filter coefficients. SYNOPSIS [wtf] = wtfilter(wtfname, transform)
INPUTS
* wtfname -- name of wavelet transform filter (string, case-insenstive). * transform -- name of wavelet transform (string, case-insenstive).
OUTPUTS
* wtf -- wavelet tranform filter struct (wtf_s).
SIDE EFFECTS
DESCRIPTION
wtfilter returns a wtf_s struct with the wavelet (high-pass) and scaling (low-pass) filter coefficients, and associated attributes. The wtf_s struct has fields: * g -- scaling (low-pass) filter coefficients (vector). * h -- wavelet (high-pass) filter coefficients (vector). * L -- filter length (= number of coefficients) (integer). * Name -- name of wavelet filter (character string). * WTFclass -- class of wavelet filters (character string). * Transform -- name of transform (character string). Typing wtfilter('list') displays a list of supported filters. Typing wtfilter('all') returns a struct array of wtf_s of all supported filters. The MODWT filter coefficients are calculated from the DWT filter coefficients: ht = h / sqrt(2) gt = g / sqrt(2) The wavelet filter coefficients (h) are calculated from the scaling filter coefficients via the QMF function (wmtsa_qmf).
USAGE
% Return a wtf struct. wtf = wtfilter(wtfname, transform) % Display and return a list of available filter ames. wtfnames = wtfilter('list') % Display and return a struct array of wtf_s for available filters. wtfs = wtfilter('all', transform)
ERRORS
WMTSA:InvalidNumArguments WMTSA:MissingRequiredArgument WMTSA:InvalidArgumentValue
EXAMPLE
wtf = wtfilter('LA8', 'modwt') wtf = wtfilter('haar', 'dwt')
ALGORITHM
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
wtf_s, wtf_qmf TOOLBOX wmtsa/dwt CATEGORY Filters: Filters
AUTHOR
Charlie Cornish
CREATION DATE
2005-02-28
COPYRIGHT
(c) Charles R. Cornish 2005
CREDITS
Based on the original function (myfilter.m) by Brandon Whitcher.
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_qmf -- Calculate quadrature mirror filter (QMF).
USAGE
[b] = function_name(a, [inverse])
INPUTS
* a -- filter coefficients (vector). * inverse -- (optional) flag for calculating inverse QMF (Boolean). Default: inverse = 0 (FALSE).
OUTPUTS
b - QMF coefficients (vector).
SIDE EFFECTS
DESCRIPTION
wmtsa_qmf calculates the quadrature mirror filter (QMF) of for the specified filter coefficients. If a is a vector, the QMF of the vector is calculated. If a is a matrix or higher order array, the QMF is calculated along the first dimension. The inverse flag, if set, calculates the inverse QMF. inverse is a Boolean values specified as (1/0, y/n, T/F or true/false).
EXAMPLE
% h is the QMF of g. g = [0.7071067811865475 0.7071067811865475]; h = wmtsa_qmf(g); % g is the inverse QMF of h. h = [0.7071067811865475 -0.7071067811865475]; g = wmtsa_qmf(h, 1);
ALGORITHM
g_l = (-1)^(l+1) * h_L-1-l h_l = (-1)^l * g_L-1-l See pages 75 of WMTSA for additional details.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
yn TOOLBOX wmtsa/dwt CATEGORY Filters: Utilities
AUTHOR
Charlie Cornish
CREATION DATE
2005-02-02
COPYRIGHT
(c) 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_norminv -- Wrapper function for norm2inv. SYNOPSIS X = wmtsa_chi2inv(p) global WMTSA_USE_QGAUSS WMTSA_USE_QGAUSS = 1 X = wmtsa_norminv(p, nu)
INPUTS
* p -- probability (lower tail) with range [0,1].
OUTPUTS
* X -- corresponding inverse normal culmulative distribution function (cdf).
SIDE EFFECTS
DESCRIPTION
wmtsa_norminv is a wrapper function for toolbox norminv function. If the MATLAB Statistics toolbox is installed and a license avaialble, the norminv Statistic toolbox norminv funciton is called. Otherwise, the WMTSA stats toolbox QGauss function is called. Setting the global variable WMTSA_USE_QGAUSS to 1 will cause the QGauss to be used regardless if the MATLAB Statistics toolbox norminv is available.
USAGE
WARNINGS
ERRORS
EXAMPLE
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
(c) 2005 Charles R. Cornish MATLAB VERSION 7.0
CREDITS
REVISION
$Revision$
[top]
NAME
wmtsa_isvector_tcase -- munit test case for wmtsa_isvector.
USAGE
run_tcase('wmtsa_isvector_tcase')
INPUTS
(none)
OUTPUTS
* tc -- tcase case struct (tcase_s)
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-25
COPYRIGHT
(c) Charles R. Cornish 2005
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_isvector -- Determine if item is a vector. SYNOPSIS [tf, nsdim] = wmtsa_isvector(x, [type])
INPUTS
* x -- item to check (object). * type -- (optional) type of vector (character string).
OUTPUTS
* tf -- flag indicating whether item is a vector (Boolean). * nsdim -- the non-singleton dimension of the vector (integer).
DESCRIPTION
Function checks if the item is a vector by determining whether it has: * two dimensions * at least one singleton dimension (length of dimension = 1). The optional input argument 'type' specifies whether to check for a particular type of vector. Valid values for type include: * 'row' -- row vector with a singleton dimension of 1 * 'col','column' -- column vector with a singleton dimension of 2 * 'nonsingleton', 'truevector'-- a vector having one non-singleton dimension, i.e. either a row or column vector. * 'point' -- the degenerate case where both dimensions are singletons. There is no default value for 'type'. If 'type' is not specified, any vector (row, column, point) returns a Boolean true. If the output argument 'nsdim' is specified, the ordinal value of the non-singleton dimension of the vector is returned. Valid values for nsdim are: * 1 -- first dimension is non-singleton, i.e. a row vector * 2 -- second dimension is non-singleton, i.e. a column vector * <empty> -- both dimensions are singleton, i.e. a 'point' vector.
USAGE
tf = wmtsa_isvector(x) tf = wmtsa_isvector(x, type) [tf, nsdim] = wmtsa_isvector(x)
ERRORS
WMTSA:InvalidNumArguments, WMTSA:InvalidArgumentValue
EXAMPLE
x = [1:10]; % A row vector tf = wmtsa_isvector(x) % Result: tf = 1 tf = wmtsa_isvector(x, 'row') % Result: tf = 1 tf = wmtsa_isvector(x, 'col') % Result: tf = 0 y = x'; % y is a column vector. tf = wmtsa_isvector(y, 'row') % Result: tf = 0 tf = wmtsa_isvector(y, 'col') % Result: tf = 1 [tf, nsdim] = wmtsa_isvector(y) % Result: tf = 1, nsdim = 1 [tf, nsdim] = wmtsa_isvector(x) % Result: tf = 1, nsdim = 2
NOTES
1. Starting with version 7, MATLAB features a isvector function. wmtsa_isvector is compatiable with MATLAB version but supplies additional functionality. TOOLBOX wmtsa/utils CATEGORY WMTSA Utilities
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-26
COPYRIGHT
(c) 2004, 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_isscalar_tcase -- munit test case for wmtsa_isscalar.
USAGE
run_tcase('wmtsa_isscalar_tcase')
INPUTS
(none)
OUTPUTS
* tc -- tcase case struct (tcase_s)
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-25
COPYRIGHT
(c) Charles R. Cornish 2005
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_isscalar -- Determine if item is a scalar.
USAGE
[tf] = isscalar(x)
INPUTS
* x -- item to check (object).
OUTPUTS
* tf -- flag indicating whether item is a point (numeric Boolean).
DESCRIPTION
In MATLAB, points, vectors and matrices all have a dimensionality of two, (i.e. ndims(x) = 2). A point is the degenerate case where an array, has size of one in all dimensions, i.e. the array is singleton in all dimensions. Function checks whether item is a point by checking that its length in all dimensions is one.
ERRORS
WMTSA:InvalidNumArguments TOOLBOX wmtsa/utils CATEGORY WMTSA Utilities
AUTHOR
Charlie Cornish
CREATION DATE
2005-Feb-03
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_ismatrix_tcase -- munit test case for wmtsa_ismatrix.
USAGE
run_tcase('wmtsa_ismatrix_tcase')
INPUTS
(none)
OUTPUTS
* tc -- tcase case struct (tcase_s)
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-25
COPYRIGHT
(c) Charles R. Cornish 2005
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_ismatrix -- Determine if item is a matrix.
USAGE
[tf] = wmtsa_ismatrix(x, [type])
INPUTS
* x -- item to check (object). * type -- (optional) type of matrix (character string). Valid Values: 'point', 'truevector', 'truematrix' Default: No type specified.
OUTPUTS
* tf -- flag indicating whether item is a vector (Boolean). * sz -- size of x (vector of length 2).
DESCRIPTION
By definition a matrix is a two dimensional array, which degenerates to a vector or a point if it has, respectively, one or two singleton dimensions. In MATLAB, points, vectors and matrices all have a dimensionality of two, (i.e. ndims(x) = 2). A vector is the case of where one dimension has length equal to one. A point is the degenerate case where both dimensions have lengths equal to one. The function checks whether item * item has 2 dimensions If type is specified, it checks whether x is: * a 'point' having two singleton dimensions. * a 'truevector' having one singleton dimensions. * a 'truematrix' having no singleton dimensions. If the output argument 'sz' is specified, the size of the dimensions of the item are returned.
USAGE
tf = wmtsa_ismatrix(x) tf = wmtsa_ismatrix(x, type) [tf, sz] = wmtsa_ismatrix(x)
ERRORS
WMTSA:InvalidNumArguments, WMTSA:InvalidArgumentValue
EXAMPLE
x = ones(10,10); % A matrix tf = wmtsa_ismatrix(x) % Result: tf = 1 tf = wmtsa_ismatrix(x, 'point') % Result: tf = 0 tf = wmtsa_ismatrix(x, 'truevector') % Result: tf = 0 [tf, sz] = wmtsa_ismatrix(x) % Result: tf = 1, sz = [10 10] TOOLBOX wmtsa/utils CATEGORY WMTSA Utilities
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-26
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_gen_fd_sdf_acvs -- Generate the ACVS from the SDF of fractionally difference (FD) process. SYNOPSIS s_X = wmtsa_gen_fd_sdf_acvs(N, delta, sigma_squared)
INPUTS
* N -- number of data points in series (integer). * delta -- long memory parameter for FD process (vector). * sigma_squared -- (optional) process variance. Default: 1
OUTPUTS
* s_X -- ACVS of SDF of FD process (2*N+1 x length(delta) array).
SIDE EFFECTS
DESCRIPTION
For given delta(s) and series length N, function calculates the autocovariance sequence (ACVS) of the spectral density function (SDF) of a fractionally differenced (FD) process. delta may be vector of values in the range -1 =< delta < 0.5.
USAGE
WARNINGS
ERRORS
EXAMPLE
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_encode_errmsg -- Encode error message for given err_id. SYNOPSIS [errmsg] = wmtsa_encode_errmsg(err_id, [varargin])
INPUTS
err_id = error id. varagin = variable argument list.
OUTPUTS
errmsg = error message.
SIDE EFFECTS
DESCRIPTION
Function encodes an error message for a specified err_id using the variable number of arguments passed on the funtion call.
EXAMPLE
WARNINGS
ERRORS
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
2004-06-25
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_data_tcase -- munit test case to test wmtsa_data.
USAGE
run_tcase('wmtsa_data_tcase')
INPUTS
OUTPUTS
tc = tcase structure for wmtsa_data testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
wmtsa_data
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_data -- Load a WMTSA dataset. SYNOPSIS wmtsa_data(dataset) [X, att] = wmtsa_data(dataset) [dataset_list] = wmtsa_data('datasets')
INPUTS
* dataset -- name of sample dataset (string).
OUTPUTS
* out -- listing of available datasets (cell array of strings).
SIDE EFFECTS
DESCRIPTION
wmtsa_data loads a dataset into the caller's workspace. Two variables are created in the caller's workspace: * dataset -- variable containing the data * dataset_att -- attributes of dataset. A function named 'load_<dataset>' exists for each dataset and contains the instructions on how to load the dataset and create its attributes. The load functions reside in the same directory (toolbox) as the wmtsa_data function. Calling wmtsa_data('datasets') displays a list of available datasets and returns a list to the output argument.
USAGE
wmtsa_data('ecg') % Creates ecg and ecg_att variables in workspace. [X, att] = wmtsa_data('ecg') % Creates X and att variables in workspace. wmtsa_data('ecg_la8_modwt') % Creates ecg MODWT coefficients % (ecg_WJt, ecg_VJ0t, ecg_WJt_att) [WJt, VJ0t, att] = wmtsa_data('ecg_la8_modwt') % Creates WJt, VJ0t and att variables in workspace.
EXAMPLE
wmtsa_data('ecg');
WARNINGS
ERRORS
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
2005-01-26
COPYRIGHT
(c) 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_chi2inv -- Wrapper function for chi2inv SYNOPSIS X = wmtsa_chi2inv(p, nu) global WMTSA_USE_QCHISQ WMTSA_USE_QCHISQ = 1 X = wmtsa_chi2inv(p, nu)
INPUTS
* p -- probability (lower tail) with range [0,1]. * nu -- degrees of freedom
OUTPUTS
* X -- corresponding inverse of chi-square cdf.
SIDE EFFECTS
DESCRIPTION
wmtsa_chi2inv is a wrapper function for toolbox chi2inv function. If the MATLAB Statistics toolbox is installed and a license avaialble, the chi2inv Statistic toolbox norminv funciton is called. Otherwise, the WMTSA stats toolbox QGauss function is called. Setting the global variable WMTSA_USE_QCHISQ to 1 will cause the QChisq to be used regardless if the MATLAB Statistics toolbox chi2inv is available.
USAGE
WARNINGS
ERRORS
EXAMPLE
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
(c) 2005 Charles R. Cornish MATLAB VERSION 7.0
CREDITS
REVISION
$Revision$
[top]
NAME
wmtsa_ccvs_tcase -- munit test case to test wmtsa_ccvs.
USAGE
run_tcase('wmtsa_ccvs_tcase')
INPUTS
OUTPUTS
tc = tcase structure for wmtsa_ccvs testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
wmtsa_ccvs
AUTHOR
Charlie Cornish
CREATION DATE
2004-06-25
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_ccvs -- Calculate the cross covariance sequence (CCVS) of a data series. SYNOPSIS CCVS = wmtsa_ccvs(X, [dim], [estimator], [subtract_mean], [method])
INPUTS
* X -- set of observations (matrix or vector). * Y -- set of observations (matrix or vector). * dim -- (optional) dimension to calculate CCVS over (integer). * estimator -- (optional) type of estimator Valid values: 'biased', 'unbiased', 'none' Default: 'biased' * subtract_mean -- (optional) flag whether to subtract mean (numeric Boolean). Default: 1 = subtract mean * method -- method used to calculate CCVS (character string).
OUTPUTS
* CCVS -- crossvariance sequence (CCVS) (vector of length N).
SIDE EFFECTS
X, Y are real; otherwise error. X, Y are vectors or matrices; otherwise error.
DESCRIPTION
wmtsa_ccvs calculates the crosscovariance sequence (CCVS) for a real valued series. By default, the function calculates the CCVS over the first non-singleton dimension. For the current implementation X and Y must be a vectors or matrices; higher order arrays not handled. If X and Y are a vectors, CCVS is is returned with dimensions as X. If X and Y are matrices, CCVS is calculated for the columns. If input argument 'dim' is supplied, the CCVS is calculated over that dim. The estimator option normalizes the CCVS estimate as follows: * 'biased' -- divide by N * 'unbiased' -- divide by N - tau * 'none' -- unnormalized. The 'subtract_mean' input argument specifies whether to subtract the mean from the prior to calculating the CCVS. The default is 1 = 'subtract mean'. The 'method' input argument specifies the method used to calculate the CCVS: * 'lag' -- Calculate taking lag products. * 'fft' -- Calculate via FFT. The default is 'fft'.
ERRORS
WMTSA:InvalidNumArguments, WMTSA:InvalidArgumentDataType, WMTSA:InvalidArgumentValue
EXAMPLE
ALGORITHM
See page 266 of WMTSA for definition of CCVS. See page 269 of WMTSA for definition of biased CCVS estimator.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
wmtsa_acvs
AUTHOR
Charlie Cornish Brandon Whitcher
CREATION DATE
2003-04-23 CREDIT Based on original function myACF.m by Brandon Whitcher.
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_biased_ccvs -- Calculate the biased cross covariance sequence (CCVS) of two data series. SYNOPSIS CCVS = wmtsa_biased_ccvs(X, Y)
INPUTS
X - vector of observations. Y - vector of observations.
OUTPUTS
CCVS - cross covariance sequence (CCVS) of X and Y.
DESCRIPTION
EXAMPLE
ALGORITHM
See page 269 of WMTSA.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003-04-23 Credits: Based on original function myACF.m by Brandon Whitcher.
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_acvs_tcase -- munit test case to test wmtsa_acvs.
USAGE
run_tcase('wmtsa_acvs_tcase')
INPUTS
OUTPUTS
tc = tcase structure for wmtsa_acvs testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
wmtsa_acvs
AUTHOR
Charlie Cornish
CREATION DATE
2004-06-25
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
wmtsa_acvs -- Calculate the autocovariance sequence (ACVS) of a data series. SYNOPSIS ACVS = wmtsa_acvs(X, [dim], [estimator], [subtract_mean], [method], [dim])
INPUTS
* X -- set of observations (array). * estimator -- (optional) type of estimator Valid values: 'biased', 'unbiased', 'none' Default: 'biased' * subtract_mean -- (optional) flag whether to subtract mean (numeric Boolean). Default: 1 = subtract mean * method -- (optional) method used to calculate ACVS (character string). * dim -- (optional) dimension to compute ACVS along (integer). Default: 1 or first non-singular dimension.
OUTPUTS
* ACVS -- autocovariance sequence (ACVS) (vector or matrix).
SIDE EFFECTS
X is a real; otherwise error. X is a vector or matrix; otherwise error.
DESCRIPTION
wmtsa_acvs calculates the autocovariance sequence (ACVS) for a real valued series. By default, the function calculates the ACVS over the first non-singleton dimension. For the current implementation X must be a vector or matrix, higher order arrays not handled. If X is a vector, ACVS is returned with dimensions as X. If X is a matrix, ACVS is calculated for the columns. If input argument 'dim' is supplied, the ACVS is calculated over that dim. The estimator option normalizes the ACVS estimate as follows: * 'biased' -- divide by N * 'unbiased' -- divide by N - tau * 'none' -- unnormalized. The 'subtract_mean' input argument specifies whether to subtract the mean from the prior to calculating the ACVS. The default is to subtract them mean. The 'method' input argument specifies the method used to calculate the ACVS: * 'lag' -- Calculate taking lag products. * 'fft' -- Calculate via FFT. * 'xcov' -- Calculate via xcov function. The default is 'fft'.
ERRORS
WMTSA:InvalidNumArguments, WMTSA:InvalidArgumentDataType, WMTSA:InvalidArgumentValue
EXAMPLE
ALGORITHM
See page 266 of WMTSA for definition of ACVS. See page 269 of WMTSA for definition of biased ACVS estimator.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
AUTHOR
Charlie Cornish Brandon Whitcher
CREATION DATE
2003-04-23 CREDIT Based on original function myACF.m by Brandon Whitcher.
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
wavelet_filter -- Compute wavelet filter coefficents from scaling filter coefficients. SYNOPSIS h = wavelet_filter(g)
INPUTS
g - vector of scaling filter coefficeints
OUTPUTS
h - vector of wavelet filter coefficeints
DESCRIPTION
ALGORITHM
See equation 75a of WMTSA.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003-10-01
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
verify_datatype_tcase -- munit test case to test verify_datatype.
USAGE
run_tcase(@verify_datatype_tcase)
INPUTS
OUTPUTS
tc = tcase structure for verify_datatype testcase.
SIDE EFFECTS
DESCRIPTION
AUTHOR
Charlie Cornish
CREATION DATE
2005-07-18
COPYRIGHT
(c) 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
verify_datatype -- Verify the datatype(s) of a variable. SYNOPSIS [tf, msg] = verify_datatype(var, datatypes, var_name)
INPUTS
* var -- variable to verify (object) * datatypes -- expected data type(s) of the arg. See verify_datatypes function for possibles datatypes to check (string or cell array of strings). * var_name -- (optional) alternative name of variable to use in error message (string).
OUTPUTS
* tf -- flag indicating whether object as specified datatype(s). (Boolean). * msg -- diagnostic error message (string).
SIDE EFFECTS
Function call requires a mininum of two input arguments; otherwise error.
DESCRIPTION
verify_datatypes checks whether the specified object has the specified data type(s). If the datatypes argument is a cell array of strings, then the function checks whether the variable datatype matches each and all of the datatypes. The function returns a logical true if the object's datatypes match those specified by the 'datatypes' argument; otherwise a logical false is returned. Possible datatypes to check include: * 'posint' -- All are positive integers --> integer value(s) > 0. * 'int0' -- All are positive integers plus zero --> integer value(s) >= 0. * 'int','integer' -- All are integers --> any integer value(s). * 'real' -- All are real numbers. * 'num','numeric' -- All are numeric. * 'struct','structure' -- Is a structure. * 'char','character','string' - Is a character string. * 'scalar -- Is a point (size of all dimensions = 1). * 'vec','vector' -- Is a vector (i.e. MxN, with M and/or N = 1). * 'matrix' -- Is a matrix (i.e. dim = 2). * 'nonsingleton', 'truevector' -- Is a vector (i.e. MxN with M *or* N = 1). * 'row','rowvector' -- Is a row vector (i.e. M x 1). * 'col','columnvector' -- Is a column vector (i.e. 1 x N). * 'finite' -- All are finite. * 'nonsparse' -- Is a non-sparse matrix.
EXAMPLE
ERRORS
WMTSA:InvalidNumArguments, WMTSA:InvalidArgumentType, WMTSA:InvalidArgumentValue
NOTES
SEE ALSO
argterr TOOLBOX wmtsa/utils CATEGORY WMTSA Utilities
AUTHOR
Charlie Cornish
CREATION DATE
2005-02-17
COPYRIGHT
(c) 2005 Charles R. Cornish MATLAB VERSION 7.0
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
validate_opts -- Validate opts fieldnames. SYNOPSIS [errmsg] = validate_opts(opts, valid_opts) [errmsg] = validate_opts(opts, valid_opts, 'string') [errstruct] = validate_opts(opts, valid_opts, 'struct')
INPUTS
* opts -- name-value pairs to validate (struct). * valid_opts -- valid opts names (struct or cell array of strings).
OUTPUTS
* errmsg -- error message (string). * errstruct -- error struct with fields: message, identifier.
SIDE EFFECTS
Function call requires a minimum of 2 input arguments; otherwise error.
DESCRIPTION
validate_opts validates the fieldnames in the struct 'opts' against a list of possible option names contained in valid_opts. valid_opts may be another struct or a cell array of strings. If all fieldnames are valid, the function returns with empty output arguments. If the function encounters an opt fieldname not found in the set of valid fieldnames, it returns an error message encoding the name of invalid opt fieldname.
USAGE
WARNINGS
ERRORS
EXAMPLE
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
(c) 2005 Charles R. Cornish MATLAB VERSION 7.0
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
utils_tsuite -- munit test suite to test utils.
USAGE
INPUTS
OUTPUTS
ts = tsuite structure for utils testsuite.
SIDE EFFECTS
DESCRIPTION
AUTHOR
Charlie Cornish
CREATION DATE
2005-07-18
COPYRIGHT
(c) 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
set_opts_defaults -- Set default values for opts. SYNOPSIS [opts] = set_opts_defaults(opts, opts_defaults)
INPUTS
* in_opts -- set of name-value pairs (struct). * opts_defaults -- default values for opts (struct). * default_fields_only -- return opts struct with default fields (logical)
OUTPUTS
* out_opts -- name-value pairs with default values (struct).
SIDE EFFECTS
DESCRIPTION
USAGE
WARNINGS
ERRORS
EXAMPLE
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
(c) 2005 Charles R. Cornish MATLAB VERSION 7.0
CREDITS
REVISION
$Revision: 614 $
[top]
NAME
set_infomsg_verbosity_level -- Set and return numeric value of VERBOSE to the verbose level for informational messages.
USAGE
VERBOSITY = set_infomsg_verbosity_level(verbosity_level)
INPUTS
OUTPUTS
verbosity_level = verbosity level, integer or character value Valid Values: an integer or character string with possible values: -1, silent 0, operational, none 1, verbose 2, very, veryvebose 3, extremely, extremelyvebose Default: 0 = operational
SIDE EFFECTS
Error is raised if verbosity_level is an invalid value.
DESCRIPTION
Function sets the value of the global variable VERBOSITY to following values: 1 for verbose messaging 2 for very verbose messaging 3 for extremely verbose messaging. When infomsg is called with a verbosity_level, the message is displayed based on the value of the global variable VERBOSITY set via set_infomsg_verbosity_level or manually.
NOTES
1. Global variable VERBOSITY must be declared in the calling base or caller workspace prior to executing set_infomsg_verbosity. 2. Value of verbosity_level may be an integer or character. The function evaluates verbosity_level, determineds the integer value of verbosity_level and sets VERBOSITY.
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003/05/08
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
set_defaults_tcase -- munit test case to test set_defaults.
USAGE
run_tcase('set_defaults_tcase')
INPUTS
OUTPUTS
tc = tcase structure for set_defaults testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
set_defaults
AUTHOR
Charlie Cornish
CREATION DATE
2005-07-19
COPYRIGHT
(c) 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
set_defaults -- Create variables with default values if non-existant in workspace. SYNOPSIS set_defaults(defaults, [var_names])
INPUTS
* defaults -- strut of name-value pairs of defaults (struct). * var_names -- (optional) names of specific variables to set defaults (string or cell array of strings).
OUTPUTS
SIDE EFFECTS
Function call requires a minimum of 1 input arguments; otherwise error.
DESCRIPTION
set_defaults checks whether variable(s) exist in the caller's workspace, and, if not, creates them with the supplied default values. The default values are passed as a set of name-value pairs via the 'defaults' struct argument. If the 'var_names' argument is specified, then only those variables in var_names are created. Other all variables with the fieldnames in 'defaults' are created.
USAGE
WARNINGS
ERRORS
EXAMPLE
defaults.a = 1; defaults.b = 'abc'; defaults.c = {'xyz', 1, 2, 3}; % Example 1 - Create all variables with defaults; % Variables do not exist in caller's workspace. clear a b c set_defaults(defaults); whos a b c % Example 2 - Create those variables that do not exist in caller's workspace. clear a b c b = 2 set_defaults(defaults); whos a b c b % Example 3 - Create specific variables with defaults. % Some variables do exist in caller's workspace. clear a b c a = 9 var_names = {'a', 'b'} set_defaults(defaults); whos a b c
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
(c) 2005 Charles R. Cornish MATLAB VERSION 7.0
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
set_default_tcase -- munit test case to test set_default.
USAGE
run_tcase('set_default_tcase')
INPUTS
OUTPUTS
tc = tcase structure for set_default testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
set_default
AUTHOR
Charlie Cornish
CREATION DATE
2005-07-19
COPYRIGHT
(c) 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
set_default -- Set default value for a variable in workspace. SYNOPSIS set_default(var_name, default)
INPUTS
* var_name -- Name of variable to check and set value (string). * default -- Value to set (various datatypes).
OUTPUTS
(none)
SIDE EFFECTS
DESCRIPTION
set_defaults checks whether a variable with the name 'var_name' exists in the caller's workspace. If the variable does not exist, the function creates a variable the 'var_name' name and sets its value to 'default. If the variable does exist, the function leaves the variable value unchanged.
ERRORS
EXAMPLE
set_default('x', 1);
NOTES
set_default creates and sets the default value for a single variable. The set_defaults function creates and sets the default values for multiple variables.
SEE ALSO
set_defaults TOOLBOX wmtsa/utils CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
2005-08-01
COPYRIGHT
(c) 2005 Charles R. Cornish MATLAB VERSION 7.0
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
set_plot_prop - Set properties of given axes. SYNOPSIS [haxes] = set_axes_prop(haxes, axesProp)
INPUTS
haxes = handle of axes to set properties. axesProp = stucture containing name-value pairs of axes properties to set.
OUTPUTS
haxes = handle to axes of properties set.
DESCRIPTION
set_axes_prop takes a structure containing a set of field name-value pairs and sets property values of the named properties for the specified axes. The X/Y/ZLabel property is treated specially, since this is a property of the label (i.e. Text) object.
EXAMPLE
axesProp.XLim = [0, 10]; axesProp.XTick = [0:5:10]; axesProp.XTickLabel = axesProp.XTick; axesProp.XLabel = 'x axis'; ha = gca; set_axes_prop(ha, axesProp);
ERRORS
If field name is not a valid axes property name, error occurs.
SEE ALSO
axes, Axes Properties TOOLBOX wmtsa/plotutil CATEGORY Plotting Utilties
AUTHOR
Charlie Cornish
CREATION DATE
2004-02-03
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
rgb2colorspecname -- Look up ColorSpec name by rgb value.
USAGE
[colorspecname] = rgb2colorspecname(rgb, [format])
INPUTS
rbg - three-element row vector whose elements specify the intensities of the red, green, and blue components of the color; the intensities must be in the range [0 1]. format - (optional) format for name, either 'short' or 'long' Default: 'short'
OUTPUTS
colorspecname - name of ColorSpec in specified format
SIDE EFFECTS
rgb must be RGB value for one of eight primary colors; otherwise error.
DESCRIPTION
Funciton converts a RGB 3-element vector to its equivalent name, in either ColorSpec name in either short (default) format (single letter) or long format (complete name). Possible ColorSpec names: yellow magenta cyan red green blue white black
SEE ALSO
ColorSpec TOOLBOX wmtsa CATEGORY utils
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-05
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_wvar_tcase -- munit test case to test plot_wvar.
USAGE
run_tcase('plot_wvar_tcase')
INPUTS
OUTPUTS
tc = tcase structure for plot_wvar testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
plot_wvar
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_wvar_psd_tcase -- munit test case to test plot_wvar_psd.
USAGE
run_tcase('plot_wvar_psd_tcase')
INPUTS
OUTPUTS
tc = tcase structure for plot_wvar_psd testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
plot_wvar_psd
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_wvar_psd -- Plot band-averaged power spectral density (PSD) estimated from wavelet variance. SYNOPSIS haxes = plot_wvar_psd(f, CJ, f_band, CI_CJ, mode, LineSpec)
INPUTS
f = center frequency (geometric mean) of octave band. (Jx1 vector) CJ = estimate of average of power spectral density per octave band. (Jx1 vector) f_band = (optional) lower and upper bounds of frequency octave band, (Jx2 vector) CI_CJ = (optional) confidence interval of power spectral density estimate, (Jx2 array), lower bound (row 1) and upper bound (row 2). mode = (optional) type of plot Valid values: 'point', 'staircase', 'box' Default value: 'point' LineSpec = (optional) LineSpec string axes_scale = (optional) Type of axes scale to use Valid values: 'linear', 'loglog', 'semilogx', 'semilogy' Default value: 'loglog' axesProp = (optional) structure containing name-value pairs of axes properties to override (see axes).
OUTPUTS
haxes = handle to plot axes
SIDE EFFECTS
DESCRIPTION
plot_wvar_psd plots the MODWT estimate (CJ) of the spectral density function (SDF) average on a log-log axes. The function plots the SDF for the follwing modes: - 'point' = plots as points the average value of the SDF at the geometric mean of the octave frequency band. - 'line' = plots as a line the average value of the SDF at the geometric mean of the octave frequency band. - 'staircase' = plots the average value of SDF as a series of staircase with step width equal to width of the octave frequency band. - 'box' = plots an uncertainity box whose dimensions are the octave frequency band (width) and confidence interval of CJ.
EXAMPLE
[WJt, VJ0t] = modwt(X, 'la8', 10, 'reflection'); [wvar, CI_wvar] = modwt_wvar(WJt); [f, CJ, f_band, CI_CJ] = modwt_wvar_sdf(wvar, delta_t, CI_wvar); plot_wvar_psd(f, CJ, '', '', 'point'); hold on; plot_wvar_psd(f, CJ, f_band, '', 'staircase'); hold on; plot_wvar_psd(f, CJ, f_band, CI_CJ, 'box');
WARNINGS
ERRORS
NOTES
1. For 'staircase' mode, f_band is a required input argument. 2. For 'box' mode, f_band and CI_CJ are required input arguments.
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003-11-12
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_wvar - Plot the wavelet variance and confidence intervals. SYNOPSIS [haxes] = plot_wvar(wvar, [CI_wvar], [title_str], [ylabel_str], [LineSpec], ... [axesProp], [level_range], [plotOpts])
INPUTS
wvar = wavelet variance(s) (Jx1 vector of cell array of vectors). CI_wvar = (optional) confidence interval of wavelet variance, (Jx2 array or a cell array of Jx2 arrays) lower bound (column 1) and upper bound (column 2). title_str = (optional) character string containing title for the plot. ylabel_str = (optional) character string containing first line of ylabel for plot. LineSpec = (optional) character string containing LineSpec (see plot). axesProp = (optional) structure containing name-value pairs of axes properties to override (see axes). level_range = (optional) number or range of numbers indicating subset of levels (j's) to plot. plotOpts = (optional) structure containing plot options.
OUTPUTS
haxes = handle to axes of wavelet variance plot.
SIDE EFFECTS
DESCRIPTION
plot_wvar plots the wavelet variance and optionally confidence intervals on a semilog y-axes as a function of level. By default, the wavelet variance is plotted for range of levels from 0 to max(level) rounded up to next even number. Use axesProp.XLim to change the limits for range of levels to plot. The argument plotOpts is structure with fields controlling plotting options as follows: LabelScales = boolean flag indicating whether to add a second set of labels at min and max x-axis values with physical scale based on value of DeltaT. A second x-axis label titled scale is added to the x-axis. DeltaT = sampling interval of time series in physical units. DeltaTUnits = units of sample unit DeltaT. If specified, units is is added to second x-axis label for the scale labels. LegendList = character cell array to add a legend to plot using the values contained in the cell array. LegendPos = integer indicating where to place legend on plot. See legend for possible values and default. LineSpecList = cell array of LineSpec values to use for plotting wvar and confidence intervals. PlotErrorBar = boolean flag indicating to plot confidence intervals as error bars. EXAMPLES [WJ0t, VJ0t] = modwt(X, 'la8', 6, 'reflection') [wvar, CI_wvar] = modwt_wvar(WJt) axesProp.XLim = [0, 10] plotOpts.LabelScales = 1 plotOpts.DeltaT = (1 / 25) * 4 % Sample rate of 25 Hz times 4 m/s sensor % velocity plotOpts.DeltaTUnits = 'm' plot_wvar(wvar, CI_wvar, 'Wavelet Variance', 'w', '', axesProp, '', plotOpts)
NOTES
The function expects that complete vectors of wavelet varianaces and confidence levels are passed in as arguments. To plot a subset of points to plot, use the level_range argument.
BUGS
When first used in a new figure window, an extra line of spacing occurs between x-ticklabels and xlabel for primary and secondary x-axis labels. Reissuing the command in the same figure window will create appropriate line spacing.
SEE ALSO
modwt_wvar, plot, LineSpec, axes TOOLBOX wmtsa/plotutils CATEGORY WMTSA Plotting: Wavelet Variance
AUTHOR
Charlie Cornish
CREATION DATE
2003/04/23
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_wcov -- Plot the wavelet covariance and confidence intervals. SYNOPSIS [haxes] = plot_wvcov(wcov, [CI_wvar], [title_str], [ylabel_str], [LineSpec], ... [axesProp], [level_range], [plotOpts])
INPUTS
wcov = vector containing wavelet covariance for J levels. CI_wcov = (optional) Jx2 matrix containing confidence intervals of wcov, lower bound (column 1) and upper bound (column 2) title_str = (optional) character string containing title for the plot. ylabel_str = (optional) character string containing first line of ylabel for plot. LineSpec = (optional) character string containing LineSpec (see plot). axesProp = (optional) structure containing name-value pairs of axes properties to override (see axes). level_range = (optional) number or range of numbers indicating subset of levels (j's) to plot. plotOpts = (optional) structure containing plot options (unused at present).
OUTPUTS
haxes = handle to axes of wavelet covariance plot.
SIDE EFFECTS
DESCRIPTION
plot_wcov plots the wavelet covariance and optionally confidence intervals on a semilog y-axes as a function of level. By default, the wavelet variance is plotted for range of levels from 0 to max(level) rounded up to next even number. Use axesProp.XLim to change the limits for range of levels to plot.
EXAMPLE
[WJ0t_x, VJ0t_x] = modwt(X, 'la8', 6, 'reflection') [WJ0t_y, VJ0t_y] = modwt(Y, 'la8', 6, 'reflection') [wcov, CI_wcov] = modwt_wcov(WJt_x, WJt_y) axesProp.XLim = [0, 10] plot_wvar(wvar, CI_wvar, 'Wavelet Variance', 'w', '', axesProp)
NOTES
SEE ALSO
modwt_wcov TOOLBOX wmtsa/plotutils CATEGORY WMTSA Plotting: Wavelet Coariance
AUTHOR
Charlie Cornish
CREATION DATE
2003/04/24
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_wcor -- Plot the wavelet correlation and (optionally) confidence intervals. SYNOPSIS plot_wcor(wcor, [CI], [ylabel_str], [title_str])
INPUTS
wcor = vector containing wavelet corrrelation for J levels. CI = Jx2 matrix containing 95% confidence intervals, lower bound (column 1) and upper bound (column 2) ylabel_str = string containing first line of ylabel for plot title_str = string containing title for the plot
OUTPUTS
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
REFERENCES
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003/04/24
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_modwt_wvar_ci_comparison -- Plot MODWT wavelet variance and confidence intervals calculated via different methods. SYNOPSIS plot_modwt_wvar_ci_comparison(WJt, [ylabel_str], [title_str])
INPUTS
WJt = NxJ matrix containing MODWT-computed wavelet coefficients where N = number of time intervals, J = number of levels ylabel_str = string containing first line of ylabel for plot title_str = string containing title for the plot
OUTPUTS
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
REFERENCES
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003/05/24
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_modwt_rcwsvar -- Plot the rotated cumulative sample variance of the MODWT wavelet coefficients.
USAGE
[haxes] = plot_modwt_rcwsvar(rcwsvar, [title_str], [xaxis], [xlabel_str], ... [axesProp], [level_range])
INPUTS
rcwsvar = NxJ containing rotated cumulative sample variance of the MODWT wavelet coefficients title_str = (optional) character string or cell array of strings containing title of plot. xaxis = (optional) vector of values to use for plotting x-axis. xlabel_str = (optional) character string or cell array of strings containing label x-axis. axesProp = (optional) structure containing axes property values to override for C subplot. level_range = (optional) number or range of numbers indicating subset of levels (j's) to plot.
OUTPUTS
haxes = (optional) handle to axes for original data series (C) subplot.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
REFERENCES
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003/05/16
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_modwt_coef2 -- Plot the MODWT wavelet and scaling coefficients and the original time series (individual subplots version). SYNOPSIS [hWplotAxes, hXplotAxes] = plot_modwt_coef2(WJt, VJ0t, [X], w_att, ... [title_str], [xaxis], [xlabel_str], ... [WplotAxesProp], [XplotAxesProp], ... [J0], [level_range], [plotOpts], ... [masterPlotFrame], [xaxis_range], [scale_str]) [hWplotAxes, hXplotAxes] = plot_modwt_coef2(WJt, [], [X], w_att, ... [title_str], [xaxis], [xlabel_str], ... [WplotAxesProp], [XplotAxesProp], ... [J0], [level_range], [plotOpts], ... [masterPlotFrame], [xaxis_range], [scale_str]) [hWplotAxes, hXplotAxes] = plot_modwt_coef2([], VJ0t, [X], w_att, ... [title_str], [xaxis], [xlabel_str], ... [WplotAxesProp], [XplotAxesProp], ... [J0], [level_range], [plotOpts], ... [masterPlotFrame], [xaxis_range], [scale_str])
INPUTS
WJt = NxJ array of MODWT wavelet coefficents where N = number of time intervals, J = number of levels VJ0t = Nx1 vector of MODWT scaling coefficients at level J0. X = (optional) vector of observations of size N. * w_att -- MODWT transform attributes (struct). title_str = (optional) character string or cell array of strings containing title of plot. xaxis = (optional) vector of values to use for plotting x-axis. xlabel_str = (optional) character string or cell array of strings containing label x-axis. WplotAxesProp = (optional) structure containing axes property values to override for W subplot. XplotAxesProp = (optional) structure containing axes property values to override for X subplot. J0 = (optional) override value of J0, if J ~= J0 or if max(level_range) ~= J0. level_range = (optional) number or range of numbers indicating subset of levels (j's) to plot. plotOpts = (optional) structure containing plot options. masterPlotFrame = (optional) structure containing coordinates to place X and W plots. xaxis_range = (optional) range of xaxis values to plot. scale_str = (optional) character cell array of strings containing physcial scale values of levels to label left y-axis of W plot.
OUTPUTS
hWplotAxes = (optional) handle to axes for MODWT coefficients (W) subplot. hXplotAxes = (optional) handle to axes for original data series (X) subplot.
SIDE EFFECTS
1. If plotting VJ0t, without WJt (i.e. WJt = []), must specify a value for J0; otherwise error. 2. If a level_range is specified, must provide a value for J0; otherwise error. 3. Either WJt or VJ0t, or both WJt and VJ0t may be specified'; otherwise error.
DESCRIPTION
plot_modwt_coef plots the MODWT coefficients and optionally the original data series, each as indivual subplots. It is similar to plot_modwt_coef but differs in that thate wavelet coefficeints at each level and scaling coefficients at J0 level are plotted on their individual plot axes. By default, the Y-axis of each level is scaled to min and max values of the coefficients at that level. Either or both the MODWT wavelet and scaling coefficients may be plotted. The MODWT coefficients are circularly shifted at each level so as to properly align the coefficients with the original data data series. By default, the wavelet coefficients (WJt) and scaling coefficient at level J0 (VJ0t) are plotted. A subrange of wavelet coefficient levels may be specified by via the parameter, level_range = [lower:upper] By default, the boundaries demarcing the circularly shifted MODWT coefficients influenced by the circularity conditions are plot. Plotting of the boundaries may be toggled off. By default, the mean value of VJ0t is subtracted from VJ0t, so that when plotting a data series with large mean offsets, the VJ0t level does not dominate the Wplot. Defaults for plotOpts: plotOpts.PlotMODWTBoundary = 1; plotOpts.PlotWJt = 1; plotOpts.PlotVJ0t = 1; plotOpts.PlotX = 1; plotOpts.SubtractMeanVJ0t = 1;
EXAMPLE
[WJt, VJ0t] = modwt(X, 'la8', 10, 'reflection'); plot_modwt_coef(WJt, VJ0t, X, 'la8');
NOTES
TODO
Plotting of MODWT Boundaries is not currently implement.
REFERENCES
SEE ALSO
plot_modwt_coef, modwt, modwt_filter, overplot_modwt_cir_shift_coef_bdry,
AUTHOR
Charlie Cornish
CREATION DATE
2004/02/17
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_modwt_coef -- Plot the MODWT wavelet and scaling coefficients and the original time series. SYNOPSIS [hWplotAxes, hXplotAxes] = plot_modwt_coef(WJt, VJ0t, [X], w_att, ... [title_str], [xaxis], [xlabel_str], ... [WplotAxesProp], [XplotAxesProp], ... [J0], [level_range], [plotOpts], ... [masterPlotFrame], [xaxis_range], [scale_str]) [hWplotAxes, hXplotAxes] = plot_modwt_coef(WJt, [], [X], w_att, ... [title_str], [xaxis], [xlabel_str], ... [WplotAxesProp], [XplotAxesProp], ... [J0], [level_range], [plotOpts], ... [masterPlotFrame], [xaxis_range], [scale_str]) [hWplotAxes, hXplotAxes] = plot_modwt_coef([], VJ0t, [X], w_att, ... [title_str], [xaxis], [xlabel_str], ... [WplotAxesProp], [XplotAxesProp], ... [J0], [level_range], [plotOpts], ... [masterPlotFrame], [xaxis_range], [scale_str])
INPUTS
WJt = NxJ array of MODWT wavelet coefficents where N = number of time intervals, J = number of levels VJ0t = Nx1 vector of MODWT scaling coefficients at level J0. X = (optional) vector of observations of size N. * w_att -- MODWT transform attributes (struct). title_str = (optional) character string or cell array of strings containing title of plot. xaxis = (optional) vector of values to use for plotting x-axis. xlabel_str = (optional) character string or cell array of strings containing label x-axis. WplotAxesProp = (optional) structure containing axes property values to override for W subplot. XplotAxesProp = (optional) structure containing axes property values to override for X subplot. J0 = (optional) override value of J0, if J ~= J0 or if max(level_range) ~= J0. level_range = (optional) number or range of numbers indicating subset of levels (j's) to plot. plotOpts = (optional) structure containing plot options. masterPlotFrame = (optional) structure containing coordinates to place X and W plots. xaxis_range = (optional) range of xaxis values to plot. scale_str = (optional) character cell array of strings containing physcial scale values of levels to label left y-axis of W plot.
OUTPUTS
hWplotAxes = (optional) handle to axes for MODWT coefficients (W) subplot. hXplotAxes = (optional) handle to axes for original data series (X) subplot.
SIDE EFFECTS
1. If plotting VJ0t, without WJt (i.e. WJt = []), must specify a value for J0; otherwise error. 2. If a level_range is specified, must provide a value for J0; otherwise error. 3. Either WJt or VJ0t, or both WJt and VJ0t may be specified'; otherwise error.
DESCRIPTION
plot_modwt_coef plots the MODWT coefficients and optionally the original data series. Either or both the MODWT wavelet and scaling coefficients may be plotted (aka Wplot). The orignal time series is tagged the Xplot. The MODWT coefficients are circularly shifted at each level so as to properly align the coefficients with the original data series. By default, the wavelet coefficients (WJt) and scaling coefficient at level J0 (VJ0t) are plotted. A subrange of wavelet coefficient levels may be specified by via the parameter, level_range = [lower:upper] By default, the boundaries demarcing the circularly shifted MODWT coefficients influenced by the circularity conditions are plot. Plotting of the boundaries may be toggled off. By default, the mean value of VJ0t is subtracted from VJ0t, so that when plotting a data series with large mean offsets, the VJ0t level does not dominate the Wplot. Defaults for plotOpts: plotOpts.PlotMODWTBoundary = 1; plotOpts.PlotWJt = 1; plotOpts.PlotVJ0t = 1; plotOpts.PlotX = 1; plotOpts.SubtractMeanVJ0t = 1;
EXAMPLE
[WJt, VJ0t] = modwt(X, 'la8', 10, 'reflection'); plot_modwt_coef(WJt, VJ0t, X, 'la8');
NOTES
BUGS
1. MODWT Boundary lines do not draw across Y-axis range for values of VJ0t with large offsets (means).
REFERENCES
See figure 183 of WMTSA.
SEE ALSO
modwt, modwt_filter, overplot_modwt_cir_shift_coef_bdry, multi_yoffset_plot
AUTHOR
Charlie Cornish
CREATION DATE
2003/05/01
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_modwt_clcwsvar -- Plot the cumulative level of cumulative sample variance of MODWT wavelet coefficients. SYNOPSIS plot_modwt_clcwsvar(clcwsvar, [title_str], [xaxis], [xlabel_str])
INPUTS
clcwsvar = cumulative level cumulative sample wavlet variance. title_str = (optional) character string containing title of plot.
OUTPUTS
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
REFERENCES
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003/06/03
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_imodwt_mra -- Plot the inverse MODWT multiresolution analysis detail and smooth coefficients and original time series. SYNOPSIS [hDplotAxes, hXplotAxes] = plot_imodwt_mra(DJt, SJ0t, [X], mra_att, ... [title_str], [xaxis], [xlabel_str], ... [MRAplotAxesProp], [XplotAxesProp], ... [J0], [level_range], [plotOpts], ... [masterPlotFrame], [xaxis_range]) [hDplotAxes, hXplotAxes] = plot_imodwt_mra(DJt, [], [X], mra_att, ... [title_str], [xaxis], [xlabel_str], ... [MRAplotAxesProp], [XplotAxesProp], ... [J0], [level_range], [plotOpts], ... [masterPlotFrame], [xaxis_range]) [hDplotAxes, hXplotAxes] = plot_imodwt_mra( [], SJ0t, [X], mra_att, ... [title_str], [xaxis], [xlabel_str], ... [MRAplotAxesProp], [XplotAxesProp], ... [J0], [level_range], [plotOpts], ... [masterPlotFrame], [xaxis_range])
INPUTS
DJt = NxJ array of reconstituted detailed data series. where N = number of time intervals, J = number of levels SJO = Nx1 vector of reconstituted smoothed data series. X = (optional) vector of observations of size N. * mra_att -- MODWT transform attributes (struct). title_str = (optional) character string or cell array of strings containing title of plot. xaxis = (optional) vector of values to use for plotting x-axis. xlabel_str = (optional) character string or cell array of strings containing label x-axis. WplotAxesProp = (optional) structure containing axes property values to override for W subplot. XplotAxesProp = (optional) structure containing axes property values to override for X subplot. J0 = (optional) override value of J0, if J ~= J0 or if max(level_range) ~= J0. level_range = (optional) number or range of numbers indicating subset of levels (j's) to plot. plotOpts = (optional) structure containing plot options. masterPlotFrame = (optional) structure containing coordinates to place X and W plots. xaxis_range = (optional) range of xaxis values to plot.
OUTPUTS
hMRAplotAxes = (optional) handle to axes for multiresolution analysis (MRA) subplot. hXplotAxes = (optional) handle to axes for original data series (X) subplot.
SIDE EFFECTS
DESCRIPTION
plot_imodwt_mra plots the inverse MODWT detail and smowth coefficients and optionally the original data series. Either or both the inverse MODWT detail and smooth coefficients may be plotted. By default, the detail coefficients (DJt) and scaling coefficient at level J0 (SJ0t) are plotted. A subrange of wavelet coefficient levels may be specified by via the parameter, level_range = [lower:upper]
EXAMPLE
NOTES
REFERENCES
SEE ALSO
modwt, modwt_filter, overplot_modwt_cir_shift_coef_bdry, multi_yoffset_plot
AUTHOR
Charlie Cornish
CREATION DATE
2003/05/01
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_filter_sgf -- Plot squared gain functions for specified transform, wavelet filter, levels and frequencies. SYNOPSIS [haxes, hlines] = plot_filter_sgf(transform, wtfname, nlevels, [f], [title_str]... [plotOpts])
INPUTS
transform = transform (DWT or MODWT) to plot wtfname = character string or cell array of strings containing name(s) of a supported (MO)DWT scaling filter. nlevels = maximum level of scale to plot. f = (optional) array of sinsuoidal frequencies figtitle_str = (optional) string contanining name of title of figure. plotOpts = (optional) structure containing plotting options.
OUTPUTS
haxes = vector of handles to axes drawn. hlines = vector of handles to lines drawn for equivalent filters.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003/11/02
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_equivalent_filter -- Plot equivalent wavelet filters for J levels. SYNOPSIS [haxes, hlines_filter, hlines_acw, hlines_xaxis] = ... plot_equivalent_filter(transform, wavelet, J, ... [LineSpec], [title_str], ... [axesProp], [level_range], [plotOpts])
INPUTS
transform = name of wavelet transform (character string, case-insensitve) wavelet = name of wavelet filter (character string, case-insensitve) J = number of levels (integer > 0) LineSpec = (optional) line specification for plotting. figtitle_str = (optional) title of figure (character string) axesProp = (optional) axes properties to override (struct) level_range = (optional) subset of levels (j's) to plot (numeric range) plotOpts = (optional) additional plotting options (struct)
OUTPUTS
haxes = handles of axes drawn (vector) hlines_filter = handles of lines drawn for equivalent filters (vector) hlines_xaxis = handles to lines drawn for xaxis (vector) hlines_acw = handles of lines drawn for autocorrelation widths (vector)
SIDE EFFECTS
1. transform is a valid transform; otherwise error. 2. wavelet is a valid wavelet filter; otherwise error. 3. J > 0; otherwise error.
DESCRIPTION
plot_equivalent_filter plots the equivalent wavelet and scaling filter coefficients for the specified wavelet filter, number of levels and wavelet transform. The x-axis can be scaled to relative or absolute values of width of equivalent filter ( L_j). and while the y-axis can be scaled local or overall coefficent magnitudes using the the plotOpts argument. axesProp argument applies all subplot axes and can be used to override behaviour specified by the plotOpts argument. Note: Each subplot is has the 'Tag' property set to h or g + level number and can be identified for call back to a particular subplot. Use level_range to display partial range of levels or reverse order of levels, e.g., level_range = [10:-1:5]; Plotting options are controlled by setting fields in plotOpts structure argument as follows: PlotgJ = plot scaling equivalent filter. Default = 1 (yes). PlothJ = plot wavelet equivalent filter. Default = 1 (yes). DrawXAxis = draw a line reprsenting x-axis at y = 0. Default = 1 (yes). DrawYTick = draw and label Y-Axis tick marks. Default = 0 (no). NormalizeXAxisToLj = set the limits of the X-Axis to relative scale of Lj at jth leve. Default = 1 (yes). If = 0, X-axis limits for all scales is set to the absolute scale of the equivalent width at Jth scale and equivalent filters at all scales are shifted to properly align. NormalizeYAxisToAbs = set limits of Y-axis to absolute limits. The default = 0 (no)) plots each equivalent filter normalized to its min and max. If set to 1 (yes), plot all equivalent filter to the absolute min and max value of all equivalent filters. PlotAutocorrelationWidth = plot the autocorrelation width of the equivalent filter. Default = 0 (no) EqualYAxisYLim = Plot Yaxis with equal - and + magnitued, ie. = Set abs(YMin) = abs(YMax) = max(abs(YMin), abs(YMax)). Default = 1 (yes) PadXAxis = Extend x-axis limits by PadXAxis % on each end Default = 0 (no) PadYAxis = Extend y-axis limits by PadYAxis % on each end Default = 0 (no) Stairs = Plot filter as starirs. Default = 0, (no).
EXAMPLE
plot_equivalent_filter('modwt', 'la8', 6, '', 1:4);
REFERENCES
See figures 98a and 98b of WMTSA.
SEE ALSO
dwt_equivalent_filter, modwt_equivalent_filter, LineSpec
AUTHOR
Charlie Cornish
CREATION DATE
2004-02-12
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_dwt_vector -- Plot the vector W of DWT coefficients. SYNOPSIS [haxes] = plot_dwt_vector(W, NJ)
INPUTS
W = vector of DWT coefficents. NJ = (optional) vector of length J0+1 containing number of DWT wavelet (W) and scaling (VJ0) coefficients for each level. LineSpec = (optional) line specification for plotting. title_str = (optional) character string or cell array of strings containing title of plot. axesProp = (optional) structure containing name-value pairs of axes properties to override. plotOpts = (optional) structure containing plotting options.
OUTPUTS
haxes = handle to plot axes. hlineDWT = vector containing handles to line objects plotting DWT (see stem function for details). hlineBdry = handle to line object plotting boundaries between DWT levels.
SIDE EFFECTS
DESCRIPTION
plotOpts contains the following fields for controlling plotting: labelSubvectors = a Boolean controlling whether DWT coefficient subvectors are labeled. Default value: 1 = label subvectors. subvectorLabels = vector of numbers for labeling subectors. If vector element has value equal 0, that subvector is is not labele. Default: All Wj and VJ0 subvectors are labeled.
EXAMPLE
[W, NJ] = = dwt(X, 'la8', 6, 'periodic'); N = length(W); axesProp.XLim = [0, N]; [haxes, hlineDWT, hlineBdry] = plot_dwt_vector(W, NJ);
WARNINGS
ERRORS
BUGS
1. Only plots first channel.
REFERENCES
SEE ALSO
dwt, LineSpec, axes, line, stem
AUTHOR
Charlie Cornish
CREATION DATE
2004-01-08
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
plot_cum_wcov -- Plot the cumlative wavelet covariance and (optionally) confidence intervals. SYNOPSIS plot_cum_wcov(wcov, [CI_wcov], [ylabel_str], [title_str])
INPUTS
wcov = vector containing wavelet covariance for J levels. CI_wcov = Jx2 array containing 95% confidence interval of wcov, lower bound (column 1) and upper bound (column 2) ylabel_str = string containing first line of ylabel for plot title_str = string containing title for the plot
OUTPUTS
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
REFERENCES
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003/05/16
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
parse_opts_tcase -- munit test case to test parse_opts.
USAGE
run_tcase('parse_opts_tcase')
INPUTS
OUTPUTS
tc = tcase structure for parse_opts testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
parse_opts
AUTHOR
Charlie Cornish
CREATION DATE
2005-07-20
COPYRIGHT
(c) 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
parse_opts -- Parse name-value pair options into a struct. SYNOPSIS [opts] = parse_opts(varargin)
INPUTS
* varargin -- variable input argument list.
OUTPUTS
* opts -- struct of name-value pairs.
SIDE EFFECTS
Function call requires a minimum of 2 input arguments; otherwise error.
DESCRIPTION
parse_opts parses the input arguments for name-value pairs and returns the struct 'opts' filled with name-value pairs. Input arguments must be one of following: * (1) varargin list of name/value pairs, or * (2) a single argument of type struct with name/value pairs, or * (3) a single argument even-length vector of type cell with name/value pairs.
USAGE
WARNINGS
ERRORS
EXAMPLE
% Example 1: Parse varargin opts = parse_opts('a', 1, 'b', 'xyz', 'c', {'abc', 'def', 'jkl'}); % Example 2: Parse a cell array opts_list = {'a', 1, 'b', 'xyz', 'c', {'abc', 'def', 'jkl'}}; opts2 = parse_opts(opts_list); % Example 2: Parse a struct opt3 = parse_opts(opts);
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
2005-07-19
COPYRIGHT
(c) 2005 Charles R. Cornish MATLAB VERSION 7.0
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
overplot_modwt_cir_shift_coef_bdry -- Plot an overlay of the boundaries outside of which the shifted MODWT coefficients are influenced by circularity conditions. SYNOPSIS overplot_modwt_cir_shift_coef_bdry(hWplotAxes, WJt, VJ0t, att, ... [xaxis], [J0], [level_range], [lineProp])
INPUTS
hWplotAxes = handle to axes for MODWT coefficients (W) subplot. WJt = NxJ array of MODWT wavelet coefficents where N = number of time intervals, J = number of levels VJ0t = Nx1 vector of MODWT scaling coefficients at level J0. * att -- MODWT transform attributes (struct). xaxis = (optional) vector of values to use for plotting x-axis. J0 = (optional) override value of J0, if J ~= J0 or if max(level_range) ~= J0. level_range = (optional) number or range of numbers indicating subset of levels (j's) to plot. lineProp = (optional) structure containing line property values to override default line properties.
OUTPUTS
DESCRIPTION
overplot_modwt_cir_shift_coef_bdry plots an overlay of the boundaries of the circularly shifted MODWT coefficients influenced by the circularity condtions. The overlay is plotted on the subplot specified by the handle to the hWplotAxes axes. The default values of the line objects drawn are: Color: red LineWidth: 1 The default values may be overridden via by specifying an attribute and value in the lineProp parameter, e.g. lineProp.Color = 'green';
EXAMPLE
NOTES
REFERENCES
SEE ALSO
plot_modwt_coef, multi_yoffset_plot
AUTHOR
Charlie Cornish
CREATION DATE
2003/05/24
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
overplot_imodwt_cir_mra_bdry -- Plot an overlay of the boundaries of inverse MODWT MRA influenced by circularity conditions. SYNOPSIS overplot_imodwt_cir_mra_bdry(hMRAplotAxes, WJt, VJ0t, att, ... [xaxis], [J0], [level_range], [lineProp])
INPUTS
hMRAplotAxes = handle to axes for inverse MODWT MRA subplot. DJt = NxJ array of MODWT details where N = number of time intervals, J = number of levels SJ0t = Nx1 vector of MODWT smooth at level J0. * att -- MODWT transform attributes (struct). xaxis = (optional) vector of values to use for plotting x-axis. J0 = (optional) override value of J0, if J ~= J0 or if max(level_range) ~= J0. level_range = (optional) number or range of numbers indicating subset of levels (j's) to plot. lineProp = (optional) structure containing line property values to override default line properties.
OUTPUTS
SIDE EFFECTS
DESCRIPTION
The default values of the line objects drawn are: Color: red LineWidth: 1 The default values may be overridden via by specifying an attribute and value in the lineProp parameter, e.g. lineProp.Color = 'green'; EXAMPLES
NOTES
BUGS
ALGORITHM
REFERENCES
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003/05/24
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
numsigdig_diff -- Find differences in two arrays exceedning number of significant digits.
USAGE
result = numsigdig_diff(a, b, [numsigdig], [mode])
INPUTS
* x -- array or vector of values * y -- second array or vector of values * numsigdig -- (optional) number of significant digits threshold Valid values: >= 0 Default: 0 * mode -- (optional) format of returned result Valid Values: 'details', 'summary' Default: 'details'
OUTPUT
* result -- array or number containing the result of comparison.
DESCRIPTION
numsigdig_diff compares two arrays and allows approximate equality when strict equality may not exist due to minor differences due to rounding errors. numsigdig_diff subtracts two arrays and identifies those elements whose differences exceed the given number of significant digits. The function has 2 modes of operation: details = return an array of size(a) with elements = 0, for differences between a and b < number of significant digits = a-b, for differences between a and b >= number of significant digits summary = return a number whose values = 0, for no differences > number of significant digits > 0, number of elements >= number of significant digits.
AUTHOR
Charlie Cornish
CREATION DATE
2004-07-13
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
nsigdig_diff -- Find differences in two arrays exceedning number of significant digits.
USAGE
result = nsigdig_diff(a, b, [nsdig], [mode])
INPUTS
a = first array or vector of values b = second array or vector of values nsdig = (optional) number of significant digits threshold Valid values: >= 0 Default: 0 mode = (optional) format of returned result Valid Values: 'details', 'summary' Default: 'details'
OUTPUT
result = array or number containing the result of comparison.
DESCRIPTION
nsigdig_diff compares two arrays and allows approximate equality when strict equality may not exist due to minor differences due to rounding errors. nsigdig_diff subtracts two arrays and identifies those elements whose differences exceed the given number of significant digits. The function has 2 modes of operation: details = return an array of size(a) with elements = 0, for differences between a and b < number of significant digits = a-b, for differences between a and b >= number of significant digits summary = return a number whose values = 0, for no differences > number of significant digits > 0, number of elements >= number of significant digits.
AUTHOR
Charlie Cornish
CREATION DATE
2004-07-13
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
nargerr_tcase -- munit test case to test nargerr.
USAGE
run_tcase(@nargerr_tcase)
INPUTS
OUTPUTS
tc = tcase structure for nargerr testcase.
SIDE EFFECTS
DESCRIPTION
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-27
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
nargerr -- Check number of arguments to a function. SYNOPSIS [errmsg] = nargerr(func, numargin, argin, numargout, argout, [mode], [msg]) [errmsg] = nargerr(func, numargin, argin, numargout, argout, [mode], [msg], 'string') [errstruct] = nargerr(func, numargin, argin, numargout, argout, [mode], [msg], 'struct')
INPUTS
* func -- name of checked function (string or function handle). * numargin -- number of input arguments passed during function call (integer). * exp_numargin -- expected number of input arguments. (integer or range of integers). * numargout -- number of output arguments passed during function call (integer). * exp_numargout -- expected number of output arguments. (integer or range of integers). * mode -- (optional) output display mode (integer). 0 = silent 1 = verbose (default). * msg -- (optional) message string to be displayed (string). * return_type -- (optional) type of output argument (string). 'string' = return error message string (default). 'struct' = return error message struct.
OUTPUTS
* errmsg -- error message (string). * errstruct -- error struct with fields: message, identifier.
SIDE EFFECTS
Function call requires a minimum of 5 input arguments; otherwise error.
DESCRIPTION
nargerr checks whether the number of input and output arguments of a function call each are within the valid range. If the number of input (numargin) or output (numargout) arguments are not within range of expected values for input (exp_numargin) or output (exp_numargout) arguments, respectively, the function returns an error message string or error message struct. The 'exp_numarg' arguments may be specified as a(n): * integer * range of integers * string expression for range of integers For range intergers, e.g. [1:3], the value of 0 may be used for the lower bound and Inf for the upper bound to specify no lower or upper limit, respectively. The string expression option also allows specification of no lower or upper bounds, e.g. '2:' -- A minimum of two arguments and no upper bound. ':2' -- A maximum of two arguments. 'mode' controls whether information is displayed to the command window. 'msg' is an message to be displayed to the command window, if nonempty and regardless of the value of 'mode'. This allows a user-defined customized message for display. 'return_type' specifies the format of the output argument: 'string' -- output is an error message (errmsg). 'struct' -- output is a struct (errstruct) with fields message and identifier. The 'struct' option may be used in with the error function to throw an error. In this example, error(nargerr('myfunction', 2, 1, 2, 0, '', '', 'struct')) the number of input and output arguments are outside the expected range. The nargerr function returns an error struct, which in turn causes the error function to throw an error.
ERRORS
WMTSA:nargerr:invalidInputArgumentValue WMTSA:nargerr:incorrectStringFormat MATLAB:nargchk:notEnoughInputs (thrown by nargchk) MATLAB:nargchk:tooManyInputs (thrown by nargchk) MATLAB:nargoutchk:notEnoughOutputs (thrown by nargchk) MATLAB:nargoutchk:tooManyOutputs (thrown by nargchk)
EXAMPLE
% Default execution. [errmsg] = nargerr(mfilename, nargin, [1:3], nargout, [0:2]); % No display to command window. [errmsg] = nargerr(mfilename, nargin, [1:3], nargout, [0:2], 0); % Display optional msg to command window. msg = 'Usage: myfuction a b c [d]'; [errmsg] = nargerr(mfilename, nargin, [1:3], nargout, [0:2], 0, msg); % Use strings to specify range of integers with ':' syntax [err, errmsg] = nargerr(mfilename, nargin, ':3', nargout, ''); % Throw an error if one found. error(nargerr(mfilename, 2, 1, 2, 0, '', '', 'struct'))
NOTES
1. This function requires MATLAB 7, which allows error structures as input to the error function. 2. To skip number of argument checking for input or output arguments, specify exp_numargin or exp_numargout, respectively, as empty vectors ([]) or empty strings ('').
SEE ALSO
nargchk, nargoutchk, errargn
AUTHOR
Charlie Cornish
CREATION DATE
2003-06-22
COPYRIGHT
(c) 2003, 2004, 2005 Charles R. Cornish
CREDITS
nargerr was inspired by the errargn function by M. Misiti, Y. Misiti, G. Oppenheim, J.M. Poggi which is part of the MATLAB wavelet toolkit. MATLAB VERSION 7.0
REVISION
$Revision: 612 $
[top]
NAME
multi_yoffset_plot -- Plot series of stacked plots of multiple data series. SYNOPSIS [haxis] = multi_yoffset_plot(x, y, [ylabel_str], [axesProp], [left_ylabel_str])
INPUTS
x = vector of length nrow containing values to plot in x-dimension. y = nrow x ncol matrix of ncol sets of y-values to offset plot ylabel_str = (optional) cell array of length nrow containing character strings for labeling y values. axesProp = structure containing property values for customizing plot axes.
OUTPUTS
haxes = handle to the plot axes.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
REFERENCES
SEE ALSO
plot_modwt_coef
AUTHOR
Charlie Cornish
CREATION DATE
2003/05/08
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwtjm -- Calculate jth level MODWT coefficients (MATLAB implementation). SYNOPSIS [Wtout, Vtout] = modwtjm(Vtin, ht, gt, j)
INPUTS
* Vtin -- Input series for j-1 level (i.e. MODWT scaling coefficients) * ht -- MODWT wavelet filter coefficients. * gt -- MODWT scaling filter coefficients. * j -- level (index) of scale.
OUTPUTS
* Wtout -- MODWT wavelet coefficients for jth scale. * Vtout -- MODWT scaling coefficients for jth scale.
SIDE EFFECTS
DESCRIPTION
modwtjm is an implementation in MATLAB code of the MODWT transform for the jth level, and is included in the toolkit for illustrative purposes to demonstrate the pyramid algothrim. For speed considerations, the modwt function uses the C implementation of the MODWT transform, modwtj, which linked in as a MEX function.
EXAMPLE
X = wmtsa_data('ecg'); wtf = modwt_filter('la8'); % Compute the j = 1 level coefficients for ECG time series. j = 1; [Wtout, Vtout] = modwtjm(X, wft.h, wtf.g, j);
WARNINGS
ERRORS
NOTES
BUGS
TODO
ALGORITHM
See page 177-178 of WMTSA for pyramid algorithm.
REFERENCES
SEE ALSO
modwtj, modwt, modwt_filter TOOLBOX wmtsa CATEGORY dwt
AUTHOR
Charlie Cornish
CREATION DATE
2005-01-12
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwtj -- Compute MODWT coefficients for jth level.
USAGE
[Wt_j, Vout] = modwtj(Vin, ht, gt, j);
INPUTS
Vin - Initial time series, or scaling coefficients for j-1 level. ht - MODWT avelet filter coefficients. gt - MODWT Scaling filter coefficients. j - Level of decomposition.
OUTPUTS
Wt_j - MODWT wavelet coefficients for jth level. Vout - MODWT scaling coefficients (residuals) for jth level.
DESCRIPTION
modwtj is a Mex-Function written in C, which implements the Pyramid Alogrithm of the MODWT for the jth level. It is usually used as an internal function called by modwt. To compile, type: mex modwtj.c
EXAMPLE
[Wt_j, Vout] = modwtj(Vin, ht, gt, j);
ALGORITHM
See pages 177-178 of WMTSA for description of Pyramid Algorithm for the MODWT.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt
AUTHOR
Charlie Cornish Brandon Whitcher
CREATION DATE
2003-05-01
COPYRIGHT
CREDITS
Based on the original function (modwt.c) by Brandon Whitcher.
REVISION
$Revision: 612 $
[top]
NAME
modwt_wvar_verification_tcase -- munit test case to test modwt_wvar.
USAGE
run_tcase('modwt_wvar_verification_tcase')
INPUTS
OUTPUTS
tc = tcase structure for modwt_wvar_verification testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
modwt_wvar
AUTHOR
Charlie Cornish
CREATION DATE
2004-06-24
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_wvar_psd_tcase -- munit test case to test modwt_wvar_psd.
USAGE
run_tcase('modwt_wvar_psd_tcase')
INPUTS
OUTPUTS
tc = tcase structure for modwt_wvar_functionality testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
modwt_psd
AUTHOR
Charlie Cornish
CREATION DATE
2004-07-25
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_wvar_psd -- Calculate estimate of band-averaged power spectral density from MODWT wavelet variance. SYNOPSIS [CJ, f, CI_CJ, f_band] = modwt_wvar_psd(wvar, delta_t, [CI_wav])
INPUTS
wvar = wavelet variance (Jx1 vector). delta_t = sampling interval of original time series. CI_wvar = confidence interval of wavelet variance (Jx2 array). lower bound (column 1) and upper bound (column 2).
OUTPUTS
CJ = estimate of average of power spectral density per octave band. (Jx1 vector) f = center frequency (geometric mean) of octave band. (Jx1 vector) CI_CJ = confidence interval of power spectral density estimate (Jx2 array) lower bound (column 1) and upper bound (column 2). f_band = lower and upper bounds of frequency octave band. (Jx2 vector)
SIDE EFFECTS
DESCRIPTION
modwt_wvar_psd calculates an estimate of the average spectral density function from the MODWT wavelet variance. CJ, f, CI_CJ, and f_band are ordered by increasing level j and scale tau, and hence decreasing frequency f. This is in the typical order of PSD estimates by increasing frequency f. The reverse order of CJ by increasing level j allows the ready comparision of CJ decomposed to different partial levels J0.
EXAMPLE
WARNINGS
ERRORS
NOTES
BUGS
TODO
ALGORITHM
C_j = 2^j * wvar(j) * delta_t, for j = 1:J0 See page 316 of WMTSA for further details.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
loglog_modwt_psd, modwt_wvar, modwt
AUTHOR
Charlie Cornish
CREATION DATE
2003-11-09
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_wvar_functionality_tcase -- munit test case to test modwt_wvar.
USAGE
run_tcase('modwt_wvar_functionality_tcase')
INPUTS
OUTPUTS
tc = tcase structure for modwt_wvar_functionality testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
modwt_wvar
AUTHOR
Charlie Cornish
CREATION DATE
2004-06-24
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_wvar_ci_verification_tcase -- munit test case to test modwt_wvar.
USAGE
run_tcase('modwt_wvar_ci_verification_tcase')
INPUTS
OUTPUTS
tc = tcase structure for modwt_wvar_ci_verification testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
modwt_wvar_ci
AUTHOR
Charlie Cornish
CREATION DATE
2004-06-24
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_wvar_ci_functionality_tcase -- munit test case to test modwt_wvar_ci.
USAGE
run_tcase('modwt_wvar_ci_functionality_tcase')
INPUTS
OUTPUTS
tc = tcase structure for modwt_wvar_ci_functionality testcase.
SIDE EFFECTS
DESCRIPTION
AUTHOR
Charlie Cornish
CREATION DATE
2004-06-24
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
USAGE
[CI_wvar, edof, Qeta] = modwt_wvar_ci(wvar, MJ, [ci_method], [WJt], [lbound], [ubound], [p])
INPUTS
wvar = wavelet variance (1xJ vector). MJ = number of coefficients used calculate the wavelet variance at each level (Jx1). ci_method = (optional) method for calculating confidence interval valid values: 'gaussian', 'chi2eta1', 'chi2eta3' default: 'chi2eta3' WJt = MODWT wavelet coefficients (NxJ array). where N = number of time intervals J = number of levels required for 'gaussian' and 'chi2eta1' methods. lbound = lower bound of range of WJt for calculating ACVS for each level (Jx1 vector). ubound = upper bound of range of WJt for calculating ACVS for each level (Jx1 vector). p = (optional) percentage point for chi2square distribution. default: 0.025 ==> 95% confidence interval
OUTPUTS
CI_wvar = confidence interval of wavelet variance (Jx2 array). lower bound (column 1) and upper bound (column 2). edof = equivalent degrees of freedom (Jx1 vector). Qeta = p x 100% percentage point of chi-square(eta) distribution (Jx2 array). lower bound (column 1) and upper bound (column 2). AJ = integral of squared SDF for WJt (Jx1 vector).
SIDE EFFECTS
DESCRIPTION
MJ is vector containing the number of coefficients used to calculate the wavelet variance at each level. For the unbiased estimator, MJ = MJ for j=1:J0, where MJ is the number of nonboundary MODWT coefficients at each level. For the biased estimator, MJ = N for all levels. For the weaklybiased estimator, MJ = MJ(Haar), for j=1:J0, where MJ(Haar) is the number of nonboundary MODWT coefficients for Haar filter at each level.
EXAMPLE
ERRORS
WMTSA:InvalidNumArguments WMTSA:WVAR:InvalidCIMethod
NOTES
The output argument edof (equivalent degrees of freedom) is returned for the chi2 confidence interval methods. For the gaussian method, a null value is returned for edof.
ALGORITHM
See section 8.4 of WMTSA on pages 311-315.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
wmtsa_acvs TOOLBOX wmtsa/wmtsa CATEGORY ANOVA:WVAR:MODWT
AUTHOR
Charlie Cornish
CREATION DATE
2003-04-23
CREDITS
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
INPUTS
WJt = MODWT wavelet coefficients (NxJ array). where N = number of time intervals J = number of levels ci_method = (optional) method for calculating confidence interval valid values: 'gaussian', 'chi2eta1', 'chi2eta3' default: 'chi2eta3' estimator = (optional) type of estimator valid values: 'biased', 'unbiased', 'weaklybiased' default: 'biased' wtfname = (optional) name of wavelet filter (string, case-insensitive). required for 'biased' and 'weaklybiased' estimators. p = (optional) percentage point for chi2square distribution. default: 0.025 ==> 95% confidence interval w_att
OUTPUTS
wvar = wavelet variance (Jx1 vector). CI_wvar = confidence interval of wavelet variance (Jx2 array). lower bound (column 1) and upper bound (column 2). edof = equivalent degrees of freedom (Jx1 vector). MJ = number of coefficients used calculate the wavelet variance at each level (integer). wvar_att = structure containing MODWT wavelet variance attributes
SIDE EFFECTS
DESCRIPTION
EXAMPLE
ERRORS
WMTSA:InvalidNumArguments WMTSA:WVAR:InvalidCIMethod WMTSA:WVAR:InvalidEstimator WMTSA:WaveletArgumentRequired
NOTES
MJ is vector containing the number of coefficients used to calculate the wavelet variance at each level. For the unbiased estimator, MJ = MJ for j=1:J0, where MJ is the number of nonboundary MODWT coefficients at each level. For the biased estimator, MJ = N for all levels. For the weaklybiased estimator, MJ = MJ(Haar), for j=1:J0, where MJ(Haar) is the number of nonboundary MODWT coefficients for Haar filter at each level.
ALGORITHM
See section 8.3 of WMTSA on page 306. For unbiased estimator of wavelet variance, see equation 306b. For biased estimator of wavelet variance, see equation 306c.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt_wvar_ci TOOLBOX wmtsa/wmtsa CATEGORY ANOVA:WVAR:MODWT
AUTHOR
Charlie Cornish Brandon Whitcher
CREATION DATE
2003-04-23
CREDITS
Based on original function wave_var.m by Brandon Whitcher.
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_wcov_functionality_tcase -- munit test case to test modwt_wcov.
USAGE
run_tcase('modwt_wcov_functionality_tcase')
INPUTS
OUTPUTS
tc = tcase structure for modwt_wcov_functionality testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
modwt_wcov
AUTHOR
Charlie Cornish
CREATION DATE
2004-06-24
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_wcov -- Calculate the wavelet covariance of two sets of MODWT wavelet coefficients. SYNOPSIS [wcov, CI_wcov, VARwcov] = modwt_wcov(WJtX, WJtY, [ci_method], [estimator], [wtfname], [p])
INPUTS
WJtX = MODWT wavelet coefficients for X series (NxJ array). where N = number of time intervals, J = number of levels. WJtY = MODWT wavelet coefficients for Y series (NxJ array). ci_method = (optional) method for calculating confidence interval valid values: 'gaussian', 'none' default: 'gaussian' estimator = (optional) type of estimator valid values: 'biased', 'unbiased', 'weaklybiased' default: 'biased' wtfname = (optional) name of wavelet filter (string, case-insensitive). required for 'biased' and 'weaklybiased' estimators. p = (optional) percentage point for confidence interval. default: 0.025 ==> 95% confidence interval
OUTPUTS
wcov = wavelet covariance (Jx1 vector). CI_wcov = confidence interval of wavelet covariance (Jx2 array). lower bound (column 1) and upper bound (column 2). VARwcov = variance of wcov (Jx1 vector). NJt = number of coefficients used calculate the wavelet variance at each level (integer).
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
BUGS
1. This function has not been modified for multivariate datasets.
ALGORITHM
REFERENCES
Whitcher, B., P. Guttorp and D. B. Percival (2000) Wavelet Analysis of Covariance with Application to Atmospheric Time Series, \emph{Journal of Geophysical Research}, \bold{105}, No. D11, 14,941-14,962.
SEE ALSO
modwt
AUTHOR
Charlie Cornish Brandon Whitcher
CREATION DATE
2003-04-23 Credits: Based on original function wave_cov.m by Brandon Whitcher.
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_wcor -- Calculate the wavelet correlation of two sets of MODWT wavelet coefficients. SYNOPSIS [wcor, CI_wcor] = modwt_wcor(WJtX, WJtY, [p])
INPUTS
WJtX - NxJ array containing MODWT-computed wavelet coefficients for X dataset where N = number of time intervals, J = number of levels. WJtY - NxJ matrix containing MODWT-computed wavelet coefficients for Y dataset. p - (optional) percentage point for confidence interval. default: 0.025 ==> 95% confidence interval
OUTPUTS
wcor - Jx1 vector containing wavelet correlations. CI_wcor - Jx2 array containing confidence interval of wcor, lower bound (column 1) and upper bound (column 2).
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
ALGORITHM
REFERENCES
Whitcher, B., P. Guttorp and D. B. Percival (2000) Wavelet Analysis of Covariance with Application to Atmospheric Time Series, \emph{Journal of Geophysical Research}, \bold{105}, No. D11, 14,941-14,962.
SEE ALSO
modwt
AUTHOR
Charlie Cornish Brandon Whitcher
CREATION DATE
2003-04-23 Credits: Based on original function wave_cov.m by Brandon Whitcher.
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_wavelet_transfer_function -- Calculate transfer function for frequencies f for specified MODWT wavelet filter and (optionally) jth level. SYNOPSIS Ht_j = modwt_wavelet_transfer_function(f, wtfname, [j])
INPUTS
f - vector of sinsuoidal frequency. wtfname - name of a WMSTA-supported MODWT scaling filter. j - (optional) level (index) of scale.
OUTPUTS
Ht - vector of the transfer function values for MODWT wavelet filter h at frequencies f. Ht_j - vector of the transfer function values for MODWT wavelet filter h at frequencies f at jth level.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
ALGORITHM
See page 163 of WMTSA for Ht. See page 169 of WMTSA for Ht_j.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003-10-08
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_wavelet_sgf -- Calculate squared gain function for specified MODWT scaling filter at frequencies f and at (optionally) jth level. SYNOPSIS Hst = modwt_wavelet_sgf(f, wtfname) Hst_j = modwt_wavelet_sgf(f, wtfname, j)
INPUTS
f = vector of sinsuoidal frequency. wtfname = name of a WMSTA-supported MODWT scaling filter. j = (optional) level (index) of scale.
OUTPUTS
Hst = vector of squared gain function values for MODWT wavelet filter h at frequencies f. Hst_j = vector of squared gain function values for MODWT wavelet filter h at frequencies f at jth level.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
ALGORITHM
See page 163 of WMTSA for Hst. See page 202 of WMTSA for Hst_j.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt_wavelet_transfer_function
AUTHOR
Charlie Cornish
CREATION DATE
2003-10-08
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_verification_tcase -- munit test case to verify results of modwt transform.
USAGE
run_tcase('modwt_verification_tcase')
INPUTS
OUTPUTS
tc = tcase structure for modwt testcase.
DESCRIPTION
SEE ALSO
modwt_wvar
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-25
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_utils_tsuite -- munit test suite to test modwt_utils.
USAGE
run_tsuite('modwt_utils_tsuite')
INPUTS
OUTPUTS
ts = tsuite structure for modwt_utils testsuite.
DESCRIPTION
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-29
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_tsuite -- munit test suite to test modwt and inverse transforms.
USAGE
run_tsuite('modwt_tsuite')
INPUTS
OUTPUTS
ts = tsuite structure for modwt transform testsuite.
DESCRIPTION
AUTHOR
Charlie Cornish
CREATION DATE
2004-04-30
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_scaling_transfer_function -- Calculate transfer function for frequencies f for specified MODWT scaling filter and (optionally) jth level. SYNOPSIS Gt_j = modwt_scaling_transfer_function(f, wtfnamne, [j])
INPUTS
f - vector of sinsuoidal frequency. wtfnamne - name of a WMSTA-supported MODWT scaling filter. j - (optional) level (index) of scale.
OUTPUTS
G - vector of the transfer function values for MODWT scaling filter ht at frequencies f. G_j - vector of the transfer function values for MODWT scaling filter ht at frequencies f at jth level.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
ALGORITHM
See page 163 of WMTSA for Gt. See page 169 of WMTSA for Gt_j.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003-10-08
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_scaling_sgf -- Calculate squared gain function for frequencies f for specified MODWT scaling filter and (optionally) jth level. SYNOPSIS Gst_j = modwt_scaling_sgf(f, wtfname, [j])
INPUTS
f = vector of sinsuoidal frequency. wtfname = name of a WMSTA-supported MODWT scaling filter. j = (optional) level (index) of scale.
OUTPUTS
Gst = vector of squared gain function values for MODWT scaling filter h at frequencies f. Gst_j = vector of squared gain function values for MODWT scaling filter h at frequencies f at jth level.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
ALGORITHM
See page 163 of WMTSA for Gst. See page 202 of WMTSA for Gst_j.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt_scaling_transfer_function
AUTHOR
Charlie Cornish
CREATION DATE
2003-10-08
COPYRIGHT
REVISION
$Revision: 612 $
[top]
USAGE
[rwvar, CI_rwvar] = modwt_running_wvar(TWJt, [range], [step], [span], ... [ci_method], [estimator], [wtfname], [p])
INPUTS
TWJt = NxJ array of circularly shifted (advanced) MODWT wavelet coefficents range = (optional) vector containing range of indices of translated wavelet coefficients over which to calculate wavelet variance. Default value: entire range = [1+span/2, N-span/2] step = (optional) increment at which to calculate wavelet variances. Default value: 1 If = -1, use explicit indices passed by range. span = (optional) number of points in running segment over which to calculate wavelet variance. Default value: 2^J0 ci_method = (optional) method for calculating confidence interval valid values: 'gaussian', 'chi2eta1', 'chi2eta3' Default value: 'chi2eta3' estimator = (optional) type of estimator valid values: 'biased', 'unbiased' Default value: 'biased' wtfname = (optional) string containing name of a WMTSA-supported MODWT wavelet filter. p = (optional) percentage point for chi2square distribute Default value: 0.025 ==> 95% confidence interval
OUTPUTS
rwvar = N_rm x J array containing running wavelet variances, where N_rm is number of runnning means. CI_rwvar = N_rm x J x 2 array containing confidence intervals of running wavelet variances with lower bound (column 1) and upper bound (column 2). indices = Indices of time points in original data series for which rwvar values are calculated.
SIDE EFFECTS
DESCRIPTION
Function calculates the running wavelet variance from the translated (circularly shifted) MODWT wavelet coefficients. User may specify the range and steps of time points to over which to calculate wavelet variances and number of continguous values (span) to calculate each variance. The running variance is computed for a span of values center at the particular time point.
EXAMPLE
[WJt, VJ0t] = modwt(X, 'la8', 9) [TWJt, TVJ0t] = modwt_cir_shift(WJt, VJ0t) [rwvar, CI_rwvar] = modwt_running_wvar(TWJt)
WARNINGS
ERRORS
NOTES
1. User must use circularly shift MODWT wavelet coefficients. Use modwt_cir_shift prior to calculating running wavelet variances. 2. The biased estimator (default option) should be used.
BUGS
TODO
ALGORITHM
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt_wvar, modwt_cir_shift
AUTHOR
Charlie Cornish
CREATION DATE
2003-11-10
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_running_wcov -- Calculate running wavelet covariance of translated (circularly shifted) MODWT wavelet coefficients. SYNOPSIS [rwcov, CI_rwcov] = modwt_running_wcov(TWJtX, TWJtY, [range], [step], [span], [p])
INPUTS
TWJtX - NxJ array of circularly shifted (translated) MODWT wavelet coefficents for X data series. TWJtY - NxJ array of circularly shifted (translated) MODWT wavelet coefficents for Y data series. range - (optional) vector containing range of indices of translated wavelet coefficients over which to calculate wavelet variance. Default value: entire range = [1+span/2, N-span/2] step - (optional) increment at which to calculate wavelet variances. Default value: 1 span - (optional) running segment over which to calculate wavelet variance. Default value: 2^J0 p - (optional) percentage point for chi2square distribute Default value: 0.025 ==> 95% confidence interval
OUTPUTS
rwcov - N_rm x J array containing running wavelet covariances, where N_rm is number of runnning means. CI_rwcov - N_rm x J x 2 array containing confidence intervals of running wavelet covariances with lower bound (column 1) and upper bound (column 2).
SIDE EFFECTS
DESCRIPTION
Function calculates the running wavelet covariance from the translated (circularly shifted) MODWT wavelet coefficients. User may specify the range and steps of time points to over which to calculate wavelet variances and number of continguous values (span) to calculate each variance. The running variance is computed for a span of values center at the particular time point.
EXAMPLE
[WJtX, VJ0tX] = modwt(X, 'la8', 9) [WJtY, VJ0tY] = modwt(Y, 'la8', 9) [TWJtX, TVJ0tX] = modwt_cir_shift(WJtX, VJ0tX) [TWJtY, TVJ0tY] = modwt_cir_shift(WJtY, VJ0tY) [rwcov, CI_rwcov] = modwt_running_wcov(TWJtX, TWJtY)
WARNINGS
ERRORS
NOTES
1. User must use circularly shift MODWT wavelet coefficients. Use modwt_cir_shift prior to calculating running wavelet variances. 2. The biased estimator (default option) should be used.
BUGS
TODO
ALGORITHM
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt_wcov, modwt_cir_shift
AUTHOR
Charlie Cornish
CREATION DATE
2003-11-10
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_rot_cum_wsvar -- Calculate the rotated wavelet sample variance of MODWT wavelet coefficients. SYNOPSIS [rcwsvar] = modwt_rot_cum_wav_svar(WJ, wtfname)
INPUTS
WJt - NxJ array of MODWT wavelet coefficents where N = number of time intervals J = number of levels wtfname - string containing name of a WMTSA-supported MODWT wavelet filter.
OUTPUTS
rcwsvar - NxJ containing rotated cumulative sample variance of the MODWT wavelet coefficients
SIDE EFFECTS
1. wtfname is a supported wavelet, otherwise error.
DESCRIPTION
EXAMPLE
ALGORITHM
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt_cum_wav_svar, modwt_cir_shift, modwt
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-16
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_num_nonboundary_coef -- Calculate number of nonboundary MODWT coefficients for jth level.
USAGE
MJ = modwt_num_nonboundary_coef(wtfname, N, j)
INPUTS
* wtfname -- name of wavelet transform filter (string). * N -- number of samples (integer). * j -- jth level (index) of scale or range of j levels. (integer or vector of integers).
OUTPUTS
* MJ -- number of nonboundary MODWT coefficients for specified levels (integer or Jx1 vector of integers).
SIDE EFFECTS
DESCRIPTION
N-Lj+1 can become negative for large j, so set MJ = min(MJ, 0).
EXAMPLE
NOTES
ALGORITHM
M_j = N - Lj + 1 see page 306 of WMTSA for definition.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University _roPress.
SEE ALSO
modwt_filter, equivalent_filter_width
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-24
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_mra_verification_tcase -- munit test case to verify results of modwt_mra transform.
USAGE
run_tcase('modwt_mra_verification_tcase')
INPUTS
OUTPUTS
tc = tcase structure for modwt_mra testcase.
DESCRIPTION
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-25
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_mra_functionality_tcase -- munit test case to test modwt_mra functionality.
USAGE
run_tcase('modwt_mra_functionality_tcase')
INPUTS
OUTPUTS
tc = tcase structure for modwt_mra testcase.
DESCRIPTION
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-25
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_mra -- Calculate MODWT multi-resolution details and smooths of a time series (direct method). SYNOPSIS [DJt, SJt, mra_att] = modwt(X, wtfname, nlevels, boundary)
INPUTS
* X -- set of observations (vector of length N or matrix of size N x Nchan) * wtf -- (optional) wavelet transform filter name or struct (string, case-insensitve or wtf struct). Default: 'la8' * nlevels -- (optional) maximum level J0 (integer) or method of calculating J0 (string). Valid values: integer>0 or a valid method name Default: 'conservative' * boundary -- (optional) boundary conditions to use (string) Valid values: 'circular' or 'reflection' Default: 'reflection' * opts -- (optional) Additional function options.
OUTPUT
* DJt -- MODWT details coefficents (N x J x NChan array). * SJt -- MODWT smooth coefficients (N x {1,J} x NChan vector). * mra_att -- structure containing IMODWT MRA transform attributes
SIDE EFFECTS
1. wtfname is a WMTSA-supported MODWT wavelet filter; otherwise error. 2. nlevels is an integer > 0, or is a string containing valid method for choosing J0; otherwise error.
DESCRIPTION
modwt_mra calculates the MODWT MRA detail and smooth coefficients from a set of observations in a single function call. The output parameter att is a structure with the following fields: name - name of transform (= 'MODWT') wtfname - name of MODWT wavelet filter npts - number of observations (= length(X)) J0 - number of levels boundary - boundary conditions
EXAMPLE
NOTES
ALGORITHM
See pages 177-179 of WMTSA for description of Pyramid Algorithm for the inverse MODWT multi-resolution analysis.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
imodwt_mra, imodwt_details, imodwt_smooth, imodwtj, modwt, modwt_filter
AUTHOR
Charlie Cornish
CREATION DATE
2004-04-29
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_inv_cir_shift -- Inverse circularly shift (advance) MODWT coefficients. SYNOPSIS
INPUTS
TWJt - NxJ array of circularly-shifted MODWT wavelet coefficents. where N = number of time intervals. J = number of levels. TVJ0t - Nx1 vector of circularly-shifted MODWT scaling coefficients at level J0. wtfname - name of a supported WMSTA wavelet filter. J0 - largest or partial level of MODWT.
OUTPUTS
WJt - NxJ array of MODWT wavelet coefficents. VJ0t - Nx1 vector of MODWT scaling coefficients at level J0.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
If WJ is not supplied, user must supply value of J0.
ALGORITHM
See pages 114-115 equation 114b of WMTSA.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
advance_scaling_filter, advance_wavelet_filter, advance_time_series_filter, modwt
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-24
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_functionality_tcase -- munit test case to test modwt functionality.
USAGE
run_tcase('modwt_functionality_tcase')
INPUTS
OUTPUTS
tc = tcase structure for modwt testcase.
DESCRIPTION
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-25
COPYRIGHT
(c) 2004, 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_filter_tcase -- munit test case to test modwt_filter.
USAGE
run_tcase('modwt_filter_tcase')
INPUTS
OUTPUTS
tc = tcase structure for modwt_filter testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
modwt_filter
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-30
COPYRIGHT
(c) 2004, 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_filter -- Define the MODWT filter coefficients. SYNOPSIS [wtf] = modwt_filter(wtfname)
INPUTS
* wtfname -- name of wavelet transform filter (string, case-insenstive).
OUTPUTS
* wtf -- wavelet tranform filter struct (wtf_s).
SIDE EFFECTS
wtfname is a valid wavelet filter name; otherwise error.
DESCRIPTION
modwt_filter returns a wtf_s struct containing the MODWT wavelet (high-pass) and scaling (low-pass) filter coefficients. MODWT filter. The wtf_s struct has fields: * g -- scaling (low-pass) filter coefficients (vector). * h -- wavelet (high-pass) filter coefficients (vector). * L -- filter length (= number of coefficients) (integer). * name -- name of wavelet filter (string). * wtfclass -- class of wavelet filters (string). * transform -- name of transform (string). Typing modwt_filter('list') displays a list of supported filters. Typing modwt_filter('all') returns a struct array of wtf_s of all supported filters.
ERRORS
EXAMPLE
[wtf] = modwt_filter('la8');
NOTES
modwt_filter is a wrapper function around the wtfilter function.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
wtfilter, dwt_filter TOOLBOX wmtsa/dwt CATEGORY Filters: Filters
AUTHOR
Charlie Cornish Brandon Whitcher
CREATION DATE
2003-09-18
COPYRIGHT
(c) 2004, 2005 Charles R. Cornish
CREDITS
Based on the original function (myfilter.m) by Brandon Whitcher.
REVISION
$Revision: 612 $
[top]
NAME
modwt_equivalent_filter -- Calculate MODWT equivalent filter for J levels. SYNOPSIS [hJ, gJ, LJ, name] = modwt_equivalent_filter(wtfname, J)
INPUTS
wtfname = string containing name of a WMTSA-supported wavelet filter, case-insensitve. J = number of levels (integer > 0)
OUTPUTS
hJ = 1xJ numeric cell array of containing equivalent wavelet filter coefficients. gJ = 1xJ numeric cell array of containing equivalent scaling filter coefficients. LJ = 1xJ numeric cell array containing widths of the equivalent filters (L_j).
SIDE EFFECTS
DESCRIPTION
EXAMPLE
WARNINGS
ERRORS
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2004-01-27
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_cum_wav_svar -- Calculate cumulative sample variance of MODWT wavelet coefficients. SYNOPSIS [cwsvar] = modwt_cum_wav_svar(WJt, wtfname)
INPUTS
WJt - NxJ array of MODWT wavelet coefficents where N = number of time intervals J = number of levels wtfname - string containing name of a WMTSA-supported MODWT wavelet filter.
OUTPUTS
cwsvar - cumulative wavelet sample variance.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
ALGORITHM
cwsvar(j,t) = 1/N * sum( WJt^2 subscript(j,u+nuH_j mod N)) for t = 0,N-1 at jth level For details, see page 189 of WMTSA.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt_cir_shift, modwt
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-16
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_cum_level_wvar_tcase -- munit test case to test modwt_cum_level_wvar.
USAGE
run_tcase('modwt_cum_level_wvar_tcase')
INPUTS
OUTPUTS
tc = tcase structure for modwt_cum_level_wvar testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
modwt_cum_level_wvar
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_cum_level_wcov_tcase -- munit test case to test modwt_cum_level_wcov.
USAGE
run_tcase('modwt_cum_level_wcov_tcase')
INPUTS
OUTPUTS
tc = tcase structure for modwt_cum_level_wcov testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
modwt_cum_level_wcov
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_cum_level_wcov -- Compute cumulative level wavelet covariance. SYNOPSIS [clwcov, CI_clwcov] = modwt_cum_level_wcov(wcov, VARwcov, [p])
INPUTS
wcov = wavelet covariance (Jx1 vector). VARwcov = covariance of wcov (Jx1 vector). p = (optional) percentage point for confidence interval. default: 0.025 ==> 95% confidence interval
OUTPUTS
clwcov = cumulative level wavelet covariance (Jx1 vector) CI_clwcov = confidence interval of cumulative level wavelet covariance (Jx2 array). lower bound (column 1) and upper bound (column 2).
SIDE EFFECTS
DESCRIPTION
EXAMPLE
WARNINGS
ERRORS
NOTES
BUGS
TODO
ALGORITHM
clwcov_m = sum(wcov_j) for j = 1,m VARclwcov_m = sum(VARwcov_j) for j = 1,m
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
2004-06-22
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
modwt_cir_shift_wcoef_bdry_indices - Return indices of the circularly shifted MODWT wavelet coefficients influenced by circularity conditions at jth level. SYNOPSIS [lower_boundary_indices, upper_boundary_indices] = modwt_cir_shift_wcoef_bdry_indices(wtfname, N, j)
INPUTS
wtfname - string containing name of a WMTSA-supported MODWT wavelet filter. N - number of data samples. j - level (index) of scale
OUTPUTS
lower_boundary_indices = vector containing indices of circularly shifted MODWT wavelet coefficients influenced by circularity conditions at lower boundary. upper_boundary_indices = vector containing indices of circularly shifted MODWT wavelet coefficients influenced by circularity conditions at upper boundary.
SIDE EFFECTS
1. wtfname is a WMTSA-supported wavelet, otherwise error. 2. N > 0, otherwise error. 3. j > 0, otherwise error.
DESCRIPTION
modwt_cir_shift_wcoef_bdry_indices returns the indices of the circularly shifted MODWT wavelet coefficients that are affected by the circularity conditions at the boundaries.
EXAMPLE
NOTES: 1. Matlab uses an array indexing scheme starting at 1 whereas WMTSA uses zero-based arrays. An offset of 1 is added to shift array indices values.
ALGORITHM
lower_boundary = [0, L_j - 2 - abs(nuH_j)] (equation 198b of WMSTA) upper_boundary = [N - abs(nuH_j), N - 1] (equation 198b of WMSTA)
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt_filter, advance_wavelet_filter, equivalent_filter_width
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-24
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_cir_shift_wcoef_bdry -- Calculate lower and upper bounds of MODWT wavelet coefficients affected by circular shift for J0 levels. SYNOPSIS [lower_bound, upper_bound] = modwt_cir_shift_wcoef_bdry(wtfname, N, J0)
INPUTS
wtfname - string containing name of a WMTSA-supported MODWT wavelet filter. N - number of data samples. J0 - largest or partial level of MODWT.
OUTPUTS
lower_bound - J0x1 vector containing lower bound for each jth level. upper_bound - J0x1 vector containing upper bound for each jth level.
SIDE EFFECTS
1. wavelet is a WMTSA-supported wavelet, otherwise error. 2. N > 0, otherwise error. 3. J0 > 0, otherwise error.
DESCRIPTION
EXAMPLE
NOTES
ALGORITHM
See page 198 of WMTSA.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt_filter, modwt_cir_shift_scaling_coef_bdry_indices
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-24
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_cir_shift_scoef_bdry_indices - Return indices of the MODWT scaling coefficients at boundaries affected by circularly shifting for the jth level. SYNOPSIS [lower_boundary_indices, upper_boundary_indices] = modwt_cir_shift_scoef_bdry_indices(wtfname, N, j)
INPUTS
wtfname - string containing name of a WMTSA-supported MODWT wavelet filter. N - number of data samples. j - level (index) of scale
OUTPUTS
lower_boundary_indices - vector containing indices of circularly shifted MODWT scaling coefficients at lower boundary. upper_boundary_indices - vector containing indices of circularly shifted MODWT scaling coefficients at upper boundary.
SIDE EFFECTS
1. wtfname is a WMTSA-supported wavelet, otherwise error. 2. N > 0, otherwise error. 3. j > 0, otherwise error.
DESCRIPTION
EXAMPLE
NOTES
1. MATLAB uses an array indexing scheme starting at 1 whereas WMTSA uses zero-based arrays. Use an offset to shift array indices values.
ALGORITHM
lower_boundary = [0, L_j - 2 - |nuG_j|] (equation 198c of WMSTA) upper_boundary = [N - |nuG_j|, N - 1] (equation 198c of WMSTA)
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt_filter, advance_scaling_filter
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-24
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_cir_shift_scoef_bdry -- Calculate lower and upper bounds of MODWT scaling coefficients affected by circular shifting for level J0. SYNOPSIS [lower_bound, upper_bound] = modwt_cir_shift_scoef_bdry(wtfname, N, J0)
INPUTS
wtfname - string containing name of a WMTSA-supported MODWT wavelet filter. N - number of data samples. J0 - largest or partial level of MODWT.
OUTPUTS
lower_bound - lower bound index at level J0. upper_bound - upper bound index at level J0.
SIDE EFFECTS
1. wavelet is a WMTSA-supported wavelet, otherwise error. 2. N > 0, otherwise error. 3. J0 > 0, otherwise error.
DESCRIPTION
EXAMPLE
NOTES
ALGORITHM
See page 198 of WMTSA.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt_filter, modwt_cir_shift_wavelet_coef_bdry_indices
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-24
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_cir_shift_mra_bdry_indices - Return indices of the MODWT MRA coefficients influenced by circularity conditions at jth level. SYNOPSIS [lower_boundary_indices, upper_boundary_indices] = modwt_cir_shift_mra_bdry_indices(wtfname, N, j)
INPUTS
wtfname - string containing name of a WMTSA-supported MODWT wavelet filter. N - number of data samples. j - level (index) of scale
OUTPUTS
lower_boundary_indices - vector containing indices of MRA influenced circularity conditions at lower boundary for jth level. upper_boundary_indices = vector containing indices of MRA influenced circularity conditions at upper boundary for jth level.
SIDE EFFECTS
1. wtfname is a WMTSA-supported wavelet, otherwise error. 2. N > 0, otherwise error. 3. j > 0, otherwise error.
DESCRIPTION
modwt_cir_shift_mra_bdry_indices returns the indices of the MRA reconstruction (details and J0th level smooth) that are affected by the circularity conditions at the boundaries.
EXAMPLE
NOTES
1. Matlab uses an array indexing scheme starting at 1 whereas WMTSA uses zero-based arrays. An offset of 1 is added to shift array indices values.
ALGORITHM
lower_boundary = [0, L_j - 2] (page 199 of WMSTA) upper_boundary = [N - L_j + 1, N - 1] (page 199 of WMSTA)
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt_filter, equivalent_filter_width
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-24
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_cir_shift_mra_bdry -- Calculate lower and upper bounds of MODWT MRA coefficients affected by circular shift for J0 levels. SYNOPSIS [lower_bound, upper_bound] = modwt_cir_shift_mra_bdry(wtfname, N, J0)
INPUTS
wtfname - string containing name of a WMTSA-supported MODWT wavelet filter. N - number of data samples. J0 - largest or partial level of MODWT.
OUTPUTS
lower_bound - J0x1 vector containing lower bound for each jth level. upper_bound - J0x1 vector containing upper bound for each jth level.
SIDE EFFECTS
1. wtfname is a WMTSA-supported wavelet, otherwise error. 2. N > 0, otherwise error. 3. J0 > 0, otherwise error.
DESCRIPTION
EXAMPLE
NOTES
ALGORITHM
See page 199 of WMTSA.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt_filter, modwt_cir_shift_mra_bdry_indices
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-24
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_cir_shift -- Circularly shift (advance) MODWT coefficients for alignment with original time series. SYNOPSIS [TWJt] - modwt_cir_shift(WJt, [], wtfname) [TVJ0t] - modwt_cir_shift([], VJ0t, wtfname, J0) [TWJt, TVJ0t] - modwt_cir_shift(WJt, VJ0t, wtfname)
INPUTS
WJt - NxJ array of MODWT wavelet coefficents where N = number of time intervals J = number of levels VJ0t - Nx1 vector of MODWT scaling coefficients at level J0. wtfname - name of a WMSTA-supported MODWT wavelet filter. J0 - largest or partial level of MODWT.
OUTPUTS
TWJt - NxJ array of circularly shifted (advanced) MODWT wavelet coefficents TVJ0t - Nx1 vector of circularly advanced MODWT scaling coefficients at level J0.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
If WJt is not supplied, user must supply value of J0.
ALGORITHM
See pages 114-115 equation 114b of WMTSA.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
advance_scaling_filter, advance_wavelet_filter, advance_time_series_filter, modwt
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-24
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_choose_nlevels -- Select J0 based on choice, wavelet filter and data series length.
USAGE
J0 = modwt_choose_nlevels(choice, wtfname, N)
INPUTS
* choice -- choice for method for calculating J0 (string) Valid Values: 'conservative' 'max', 'maximum' 'supermax', 'supermaximum' * wtfname -- wavelet transform filter name (string) Valid Values: see modwt_filter * N -- number of observations.
OUTPUT
* J0 -- number of levels (J0) based selection criteria.
SIDE EFFECTS
1. wtfname is a WMTSA-supported MODWT wtfname, otherwise error. 2. N > 0, otherwise error.
DESCRIPTION
EXAMPLE
J0 = modwt_choose_nlevels('convservative', 'la8', N)
ERRORS
WMTSA:MODWT:InvalidNumLevels = Invalid type/value specified for nlevels.
ALGORITHM
for 'conservative': J0 < log2( N / (L-1) + 1) for 'max', 'maximum': J0 =< log2(N) for 'supermax', 'supermaximum': J0 =< log2(1.5 * N) For further details, see page 200 of WMTSA.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt_filter
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-24
COPYRIGHT
(c) 2003, 2004, 2005 Charles R. Cornish
REVISION
$Revision: 612 $
[top]
NAME
modwt -- Compute the (partial) maximal overlap discrete wavelet transform (MODWT). SYNOPSIS [WJt, VJt, att] = modwt(X, [wtf], [nlevels], [boundary], [{opts}])
INPUTS
* X -- set of observations (vector of length NX or matrix of size NX x Nchan) * wtf -- (optional) wavelet transform filter name or struct (string, case-insensitve or wtf struct). Default: 'la8' * nlevels -- (optional) maximum level J0 (integer) or method of calculating J0 (string). Valid values: integer>0 or a valid method name Default: 'conservative' * boundary -- (optional) boundary conditions to use (string) Valid values: 'circular' or 'reflection' Default: 'reflection' * opts -- (optional) Additional function options.
OUTPUTS
* WJt -- MODWT wavelet coefficents (NW x J x NChan array). * VJt -- MODWT scaling coefficients (NW x {1,J} x NChan vector). * att -- MODWT transform attributes (struct).
SIDE EFFECTS
1. wtf is either a string containing a WMTSA-supported MODWT wavelet filter or a valid wtf struct; otherwise error. 2. nlevels is an integer > 0, or is a string containing valid method for choosing J0; otherwise error.
DESCRIPTION
modwt calculates the wavelet and scaling coefficients using the maximal overlap discrete wavelet transform (MODWT). The optional input arguments have default values: * wtf -- 'la8' filter * nlevels -- 'convservative' --> J0 < log2( N / (L-1) + 1) * boundary -- 'reflection'. Optional input arguments are specified as name-value pairs: * RetainVJ -- Boolean flag indicating whether to scaling coefficients have been retained at all levels. Values: 1 = true, all VJ retained, 0 = false, VJ retained for J0 level. Default: 0, VJ retained only at J0 level. The output argument att is a structure with the following fields: * Transform -- name of transform ('MODWT') * WTF -- name of wavelet transform filter or a wtf_s struct. * NX -- number of observations in original series (= length(X)) * NW -- number of wavelet coefficients * J0 -- number of levels of partial decompsition. * NChan -- number of channels in a multivariate dataset. * Boundary -- boundary conditions applied. * Aligned -- Boolean flag indicating whether coefficients are aligned with original series (1 = true) or not (0 = false). * RetainVJ -- Boolean flag indicating whether VJ scaling coefficients at all levels have been retained (1= true) or not (0 = false).
EXAMPLE
load_ecg; [WJt, VJt, att] = modwt(ecg, 'la8', 6, 'reflection');
WARNINGS
WMTSA:MODWT:LargeJ0 = 'MODWT JO > log2(Number of samples).'
ERRORS
WMTSA:invalidBoundary = 'Invalid Transform boundary method.'
NOTES
ALGORITHM
See pages 177-178 of WMTSA for description of Pyramid Algorithm for the MODWT.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwtj, modwt_filter, modwt_choose_nlevels, nargerr, argterr TOOLBOX wmtsa/wmtsa CATEGORY Transforms: MODWT
AUTHOR
Charlie Cornish
CREATION DATE
2003-04-23
COPYRIGHT
(c) 2003, 2004, 2005 Charles R. Cornish
CREDITS
Based on the original function (modwt_dbp.m) by Brandon Whitcher.
REVISION
$Revision: 630 $
[top]
NAME
load_datasets_tcase -- munit test case to test load_datasets.
USAGE
run_tcase('load_datasets_tcase')
INPUTS
OUTPUTS
tc = tcase structure for load_datasets testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2005-08-01
COPYRIGHT
2005-01-26
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
list2struct_tcase -- munit test case to test list2struct.
USAGE
run_tcase('list2struct_tcase')
INPUTS
OUTPUTS
tc = tcase structure for list2struct testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
list2struct
AUTHOR
Charlie Cornish
CREATION DATE
2005-07-18
COPYRIGHT
(c) 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
cell2struct -- Convert a list (cell array) of name-value pairs to a struct.
USAGE
[s] = list2struct(list)
INPUTS
* list -- list of name-value pairs (cell array)
OUTPUTS
* s -- structure fieldnames with values (struct).
DESCRIPTION
list2struct converts a list (cell array) of name-value pair entries into a structure with field names (name) and field values (value).
EXAMPLE
WARNINGS
ERRORS
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX wmtsa/utils CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
2005-07-18
COPYRIGHT
(c) 2005 Charles R. Cornish
CREDITS
MATLAB VERSION 7.0
REVISION
$Revision: 612 $
[top]
NAME
linesegment_plot -- Plot a series of line segments on current plot axes. SYNOPSIS [hline] = linesegment_plot(x1, y1, x2, y2, lineProp)
INPUTS
x1 = vector of x-axis start points. x1 = vector of y-axis start points. x1 = vector of x-axis end points. x1 = vector of y-axis start points. lineProp = (optional) struct containing line properties to override.
OUTPUTS
hline = vector containing handles to Line (segment) objects drawn.
SIDE EFFECTS
DESCRIPTION
linesegment_plot plots a series of line segments defined by (x1,x2)->(x2,y2). (x1,y1) are start points; (x2,y2) are end points.
EXAMPLE
NOTES
REFERENCES
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003/05/24
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
linesegment -- Plot a set of line segments as specified by their start and end points. SYNOPSIS [hline] = linesegment(xstart, ystart, xend, yend, [lineProp])
INPUTS
xstart = vector of x-coordinates for start of line segments. ystart = vector of y-coordinates for start of line segments. xend = vector of x-coordinates for end of line segments. yend = vector of y-coordinates for end of line segments.
OUTPUTS
SIDE EFFECTS
DESCRIPTION
EXAMPLE
WARNINGS
ERRORS
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2004-01-09
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
iswtf -- Determine if input is a valid wtf struct. SYNOPSIS [tf] = iswtf(wtf)
INPUTS
* wtf -- wavelet tranform filter struct (wtf_s).
OUTPUTS
* tf -- flag indicating whether a valid valid wtf struct (Boolean)
SIDE EFFECTS
Function call requires a minimum of 1 input arguments; otherwise error.
DESCRIPTION
iswtf determines whether the input argument is a valid wtf struct, i.e. having the wtf_s struct fields: * g -- scaling (low-pass) filter coefficients (vector). * h -- wavelet (high-pass) filter coefficients (vector). * L -- filter length (= number of coefficients) (integer). * name -- name of wavelet filter (character string). * wtfclass -- class of wavelet filters (character string). * transform -- name of transform (character string).
EXAMPLE
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX wmtsa/dwt CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
(c) 2005 Charles R. Cornish MATLAB VERSION 7.0
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
infomsg -- Display an informative message based on global VERBOSITY setting.
USAGE
infomsg(msg_str, [verbosity_level])
INPUTS
msg_str = Informational message to be displayed. verbosity_level = (optional) verbosity level, integer or character value Valid Values: an integer or character string with possible values: 0, operational 2, very, veryvebose 3, extremely, extremelyvebose Default: 1 = verbose
OUTPUTS
none
SIDE EFFECTS
1. Depending on setting of global variable VERBOSITY, msg_str is displayed to the command window. 2. Error is raised if verbosity_level is an invalid value.
DESCRIPTION
infomsg displays the informative message (msg_str) to the command windwo based on the setting of the global variable VERBOSITY.
NOTES
1. Global variable VERBOSITY must be declared and defined (set) to the desired verbosity level. Use set_infomsg_verbosity_level to set the verbosity level. 2. Value of verbosity_level may be an integer or character. The function evaluates verbosity_level, determines the integer value of verbosity_level.
SEE ALSO
set_infomsg_verbosity_level
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-28
COPYRIGHT
CREDITS
argterr is a rewrite of errargt by M. Misiti, Y. Misiti, G. Oppenheim, J.M. Poggi which is part of the MATLAB wavelet toolkit.
REVISION
$Revision: 612 $
[top]
NAME
imodwtj -- Compute inverse MODWT for jth level. SYNOPSIS Vout = imodwtj(Wt_j, Vin, ht, gt, j)
INPUTS
Wt_j - MODWT wavelet coefficients for jth level. Vin - Scaling coefficients at J0, or j-1 level. ht - MODWT avelet filter coefficients. gt - MODWT Scaling filter coefficients. j - Level of decomposition.
OUTPUTS
Vout - MODWT scaling coefficients (residuals) for jth level.
DESCRIPTION
modwtj is a Mex-Function written in C, which implements the Pyramid Alogrithm of the MODWT for the jth level. It is usually used as an internal function called by imodwt or imodwt_mra. To compile, type: mex modwtj.c
EXAMPLE
Vout = imodwtj(Wt_j, Vin, ht, gt, j);
ALGORITHM
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
imodwt, imodwt_mra
AUTHOR
Charlie Cornish Brandon Whitcher
CREATION DATE
2003-05-01
COPYRIGHT
CREDITS
Based on the original function (imodwt.c) by Brandon Whitcher.
REVISION
$Revision: 612 $
[top]
NAME
imodwt_tcase -- munit test case to test imodwt function.
USAGE
run_tcase('imodwt_tcase')
INPUTS
OUTPUTS
tc = tcase structure for imodwt testcase.
DESCRIPTION
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-25
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
imodwt_smooth -- Calculate smooths at J0 level via inverse maximal overlap discrete wavelet transform (IMODWT). SYNOPSIS [SJt, att] = imodwt_smooth(VJt, wtfname, J0)
INPUTS
VJt = Nx1 vector of MODWT scaling coefficients at J0 level. wtfname = string containing name of a WMTSA-supported MODWT wavelet filter.
OUTPUT
SJOt = vector of reconstituted smoothed data series. att = structure containing IMODWT transform attributes J0 = number of levels of partial decomposition.
SIDE EFFECTS
1. wtfname is a WMTSA-supported MODWT wavelet filter; otherwise error.
DESCRIPTION
EXAMPLE
[SJt, att] = imodwt_smooth(VJt, 'la8');
NOTES
ALGORITHM
See pages 177-179 of WMTSA for description of Pyramid Algorithm for the inverse MODWT multi-resolution analysis.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
imodwtj, imodwt, imodwt_details, modwt_filter, modwt
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-01
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
imodwt_mra_verification_tcase -- munit test case to verify results of imodwt_mra transform.
USAGE
run_tcase('imodwt_mra_verification_tcase')
INPUTS
OUTPUTS
tc = tcase structure for imodwt_mra testcase.
DESCRIPTION
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-25
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
imodwt_mra -- Calculate MODWT multi-resolution details and smooths from wavelet coefficients via IMODWT transform.
USAGE
[DJt, SJt] = imodwt_smooth(WJt, VJt, wavelet)
INPUTS
* WJt -- MODWT wavelet coefficents (N x J x NChan array). * VJt -- MODWT scaling coefficients (N x {1,J} x NChan vector). * w_att -- MODWT transform attributes (struct).
OUTPUT
* DJt -- MODWT details coefficents (N x J x NChan array). * SJt -- MODWT smooth coefficients (N x {1,J} x NChan vector). * mra_att -- structure containing IMODWT MRA transform attributes
DESCRIPTION
modwt_mra computes the multi-resolution detail and smooth coefficients fomr the MODWT wavelet and scaling coefficients.
EXAMPLE
NOTES
BUGS
TODO
1. Add support for retain DJt (RetainDJt) 2. Rewrite modwt_details. 3. Rewrite modwt_smooth.
ALGORITHM
See pages 177-179 of WMTSA for description of Pyramid Algorithm for the inverse MODWT multi-resolution analysis.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
imodwt_details, imodwt_smooth, imodwtj, modwt, modwt_filter
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-01
COPYRIGHT
REVISION
$Revision: 630 $
[top]
NAME
imodwt_details -- Calculate details via inverse maximal overlap discrete wavelet transform (IMODWT). SYNOPSIS [DJt, att] = imodwt_details(WJt, wtfname)
INPUTS
WJt - NxJ array of MODWT wavelet coefficents where N = number of time points J = number of levels. wtfname - (optional) string containing name of a WMTSA-supported MODWT wavelet filter. Valid Values: see modwt_filter
OUTPUT
DJt - NxJ array of reconstituted details of data series for J0 scales. att - structure containing IMODWT transform attributes.
SIDE EFFECTS
1. wavelet is a WMTSA-supported MODWT wavelet filter; otherwise error.
DESCRIPTION
The output parameter att is a structure with the following fields: name - name of transform (= 'MODWT') wtfname - name of MODWT wavelet filter npts - number of observations (= length(X)) J0 - number of levels boundary - boundary conditions
EXAMPLE
[DJt, att] = imodwt_details(WJt, VJ0, 'la8');
NOTES
ALGORITHM
See pages 177-179 of WMTSA for description of Pyramid Algorithm for the inverse MODWT multi-resolution analysis.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
imodwtj, imodwt, imodwt_smooth, modwt_filter, modwt
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-01
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
imodwt -- Calculate the inverse (partial) maximal overlap discrete wavelet transform (IMODWT).
USAGE
X = imodwt(WJt, VJt, att)
INPUTS
* WJt -- MODWT wavelet coefficents (N x J x NChan array). * VJt -- MODWT scaling coefficients (N x {1,J} x NChan vector). * att -- MODWT transform attributes (struct).
OUTPUT
* X -- reconstituted set of observations (vector).
DESCRIPTION
imodwt computes the reconstituted time series from the MODWT wavelet and scaling coefficients.
EXAMPLE
X = imodwt(WJt, VJt, att);
ERRORS
WMTSA:InvalidNumArguments = 'Invalid number of arguments specified in function call' WMTSA:InvalidWavelet = 'Invalid wavelet filter specified'
NOTES
1. Tests indicate othat riginal and reconstituted time series agree within a precision of 10^-11, which is larger than the 10^15 numeric precision. 2. If the MODWT was calculated using 'reflection' boundary conditions, which extends the time series, and the computed coefficients, have been truncated to the length of the original series. The inverse MODWT will yield a reconstituted time series that differs from the original. If using 'reflection' boundary conditions, compute the MODWT with the opt.TruncateCoefs = 0 option (the default) to compute IMODWT which yields and exact replica of the original.
ALGORITHM
See pages 177-179 of WMTSA for description of Pyramid Algorithm for the inverse MODWT.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
modwt, imodwtj TOOLBOX wmtsa/dwt CATEGORY modwt
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-01
COPYRIGHT
(c) 2003, 2004, 2005 Charles R. Cornish
REVISION
$Revision: 632 $
[top]
NAME
fuzzy_diff -- Find differences in two arrays exceeding fuzzy_tolerance.
USAGE
result = fuzzy_diff(a, b, [fuzzy_tol], [mode])
INPUTS
a = first array or vector of values b = second array or vector of values fuzzy_tol = (optional) minimum tolerance or threshold Valid values: >= 0 Default: 0 mode = (optional) format of returned result Valid Values: 'details', 'summary' Default: 'details'
OUTPUT
result = array or number containing the result of comparison.
DESCRIPTION
fuzzy_diff compares two arrays and allows for approximate equality when strict equality may not exist due to minor differences due to rounding errors. fuzzy_diff subtracts two arrays and identifies those elements whose differences exceed the fuzzy tolerance threshold. The function has 2 modes of operation: details = return an array of size(a) with elements = 0, for differences between a and b < fuzzy_tolerance = a-b, for differences between a and b >= fuzzy_tolerance summary = return a number whose values = 0, for no differences within fuzzy_tolerance > 0, number of elements >= fuzzy_tolerance Note: fuzzy_diff compares to arrays on an absolute threshold (fuzzy_tol). Use nsigdig_diff to compare arrays that differ by a significant number of digits.
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-01
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
flipvec_case -- munit test case for flipvec.
USAGE
run_tcase('flipvector_tcase')
INPUTS
(none)
OUTPUTS
* tc -- tcase case struct (tcase_s)
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-25
COPYRIGHT
(c) Charles R. Cornish 2004, 2005
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
flipvec -- Flip a vector. SYNOPSIS [y] = flipvec(x)
INPUTS
* x -- vector of values.
OUTPUTS
* y -- vector of values in reversed order.
DESCRIPTION
Function checks whether item is a vector (row or column), i.e. has ndim = 2 and a singleton dimension, and then flips the vector (i.e. reverse order).
ERRORS
WMTSA:InvalidNumArguments WMTSA:NotAVector
SEE ALSO
isvector TOOLBOX wmtsa/utils CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-27
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
filter_center_of_energy -- Calculate filter center of energy. SYNOPSIS [coe] = filter_center_of_energy(a)
INPUTS
a = filter coefficients for filter a.
OUTPUTS
coe = center of energy for filter a.
DESCRIPTION
EXAMPLE
WARNINGS
ERRORS
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
filter_autocorrelation_width - Calculate autocorrelation width o scaling filter. SYNOPSIS [width_a] = filter_autocorrelation_width(wtfname, j, [method])
INPUTS
wtfname = string containing name of a WMTSA-supported MODWT wavelet filter, case-insensitve j = jth level (index) of scale or a range of j levels of scales (integer or vector of integers). method = (optional) name of method to use for calculations.
OUTPUTS
width_a = filter autocorrelation width for specified levels (integer or Jx1 vector of integers).
SIDE EFFECTS
1. j > 0, otherwise error.
DESCRIPTION
The function calculates the filter autocorrelation width of scaling filters for the specified wavelet. The user may optionally specify one of two methods of calculation: 'quick' - width_a = 2^j (default) 'long' - Equation 103 of WMTSA (see ALGORTHIM section.
EXAMPLE
[width_a] = filter_autocorrelation_width('haar', 5,) [width_a] = filter_autocorrelation_width('la8', 1:10, 'quick') [width_a] = filter_autocorrelation_width('d4', [2,3,5,7], 'long')
ALGORITHM
width_a {g_j,l} = (sum(g_j,l))^2 / sum(g_j,l^2) See equation 103 of WMTSA.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
dwt_equivalent_filter, dwt_filter TOOLBOX WMTSA CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
2004-12-24
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
figure_footer -- Print the string in the footer text area of a figure. SYNOPSIS [htext] = figure_footer(str, hfigure)
INPUTS
str = string to print in footer. hfigure = (optional) handle to figure to print footer. Default: current figure
OUTPUTS
htext = handle to text object of figure footer.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
WARNINGS
ERRORS
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003/06/04
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
figure_datestamp -- Print a datestamp in footer area of a figure. SYNOPSIS [htext] = figure_datestamp(filename, hfigure)
INPUTS
filename = (optional) string containing filename. hfigure = (optional) handle to figure to print footer. Default: current figure
OUTPUTS
htext = handle to text object of figure footer.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
WARNINGS
ERRORS
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003/06/04
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
equivalent_filter_width -- Calculate width of the equivalent wavelet or scaling filter. SYNOPSIS [Lj] = equivalent_filter_width(L, j)
INPUTS
* L -- width of wavelet or scaling filter (unit scale). * j -- jth level (index) of scale or a range of j levels of scales. (integer or vector of integers).
OUTPUTS
* Lj -- equivalent width of wavelet or scaling filter for specified levels (integer or Jx1 vector of integers). SIDEEFFECTS 1. L > 0, otherwise error. 2. j > 0, otherwise error.
DESCRIPTION
Given the length of a wavelet or scaling filter, the function calculates the width of the equivalent filter a level or range of levels j for the specified base filter width L.
ALGORITHM
Lj = (2^j - 1) * (L - 1) + 1 (equation 96a of WMTSA)
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-24
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
encode_errmsg -- Encode error message for specified err_id.
USAGE
[errmsg] = encode_errmsg(err_id, err_table_ps, [varargin])
INPUTS
* err_id -- error message id (character string). * err_table_ps -- error lookup table (character string or struct). * varargin -- (optional) supplemental values to encode in errmsg.
OUTPUTS
* errmsg -- encoded error message (character string).
SIDE EFFECTS
DESCRIPTION
encode_errmsg encodes an error message (errmsg) for a specified error message id (err_id). The function loads the error message table (err_table) from the specified path, searches for the matching (err_table.err_id) entry and returns the error message template (err_table.errmsg). Based on the number of message arguments (err_table.nargs), the function encodes the errmsg using the variable number of arguments (varargin) passed on the function call. The err_table is a structure array with the following fields: * err_id -- error message id (character string). * err_msg -- error message template (character string). * nargs -- number of supplement arguments to use for encoding errmsg. The input argument 'err_table_ps' may be either a character string or a struct. If a character string, err_table_ps is full path to a function or script containing the error table struct to run and load. If a struct, then the struct passed as the value in err_table_ps argument is used.
EXAMPLE
% Error table load via function wmtsa_err_table. % Name of required argument is 'transform'. errmsg = encode_errmsg('WMTSA:missingRequiredArgument', ... wmtsa_err_table, 'transform'));
WARNINGS
ERRORS
NOTES
SEE ALSO
TOOLBOX wmtsa/utils CATEGORY WMTSA Utilities
AUTHOR
Charlie Cornish
CREATION DATE
2004-06-25
COPYRIGHT
(c) 2004, 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
dwtjm -- Calculate jth level DWT coefficients (MATLAB implementation).
USAGE
[Wout, Vout] = dwtjm(Vin, h, g, j)
INPUTS
* Vin -- Input series for j-1 level (i.e. DWT scaling coefficients) * h -- DWT wavelet filter coefficients. * g -- DWT scaling filter coefficients. * j -- level (index) of scale.
OUTPUTS
* Wout -- DWT wavelet coefficients for jth scale. * Vout -- DWT scaling coefficients for jth scale.
SIDE EFFECTS
DESCRIPTION
dwtjm is an implementation in MATLAB code of the DWT transform for the jth level, and is included in the toolkit for illustrative purposes to demonstrate the pyramid algothrim. For speed considerations, the dwt function uses the C implementation of the DWT transform, modwtj, which linked in as a MEX function.
EXAMPLE
X = wmtsa_data('ecg'); wtf = dwt_filter('haar'); % Compute the j = 1 level coefficients for ECG time series. j = 1; [Wout, Vout] = dwtjm(X, h, g, j);
NOTES
BUGS
TODO
ALGORITHM
See page 100-101 of WMTSA for DWT pyramid algorithm.
REFERENCES
SEE ALSO
dwtj, dwt, dwt_filter TOOLBOX wmtsa CATEGORY dwt
AUTHOR
Charlie Cornish
CREATION DATE
2005-Sep-03
COPYRIGHT
(c) 2005 Charles R. Cornish MATLAB VERSION 7.0
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
dwt_wavelet_transfer_function -- Calculate transfer function for frequencies f for specified DWT wavelet filter and (optionally) jth level. SYNOPSIS H_j = dwt_wavelet_transfer_function(f, wavelet, [j])
INPUTS
f - vector of sinsuoidal frequency. wtfname - name of a WMSTA-supported MODWT scaling filter. j - (optional) level (index) of scale.
OUTPUTS
H - vector of the transfer function values for DWT wavelet filter h at frequencies f. H_j - vector of the transfer function values for DWT wavelet filter h at frequencies f at jth level.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
ALGORITHM
See page 69 of WMTSA for H. See page 96 of WMTSA for H_j.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003-10-08
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
dwt_wavelet_sgf -- Calculate squared gain function for specified DWT scaling filter at frequencies f and at (optionally) jth level. SYNOPSIS Hs_j = dwt_wavelet_sgf(f, wtfname, [j])
INPUTS
f = vector of sinsuoidal frequency. wtfname = name of a WMSTA-supported MODWT scaling filter. j = (optional) level (index) of scale.
OUTPUTS
Hs = vector of squared gain function values for DWT wavelet filter h at frequencies f. Hs_j = vector of squared gain function values for DWT wavelet filter h at frequencies f at jth level.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
ALGORITHM
See page 69 of WMTSA for Hs. See page 154 of WMTSA for Hs_j.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
dwt_wavelet_transfer_function
AUTHOR
Charlie Cornish
CREATION DATE
2003-10-08
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
dwt_verification_tcase -- munit test case to verify results of dwt transform.
USAGE
run_tcase('dwt_verification_tcase')
INPUTS
OUTPUTS
tc = tcase structure for dwt testcase.
DESCRIPTION
SEE ALSO
dwt_wvar
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-25
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
NAME
w_att_s -- wavelet transform attributes struct.
DESCRIPTION
w_att_s struct has fields: * Transform -- name of transform ('MODWT') (string). * WTF -- name of wavelet transform filter or a wtf_s struct (string or struct). * NX -- number of observations in original series (= length(X)) (integer). * NW -- number of wavelet coefficients (integer). * J0 -- number of levels of partial decompsition (integer). * NChan -- number of channels in a multivariate dataset (integer). * Boundary -- boundary conditions applied (string). * Aligned -- Boolean flag indicating whether coefficients are aligned with original series (1 = true) or not (0 = false) (boolean). * RetrainVJ -- Boolean flag indicating whether VJ scaling coefficients at all levels have been retained (1= true) or not (0 = false) (boolean). Possible values for transform are: DWT, MODWT.
SEE ALSO
NAME
wtf_s -- wavelet transform filter struct.
DESCRIPTION
wtf_s struct has fields: * g -- scaling (low-pass) filter coefficients (vector). * h -- wavelet (high-pass) filter coefficients (vector). * L -- filter length (= number of coefficients) (integer). * Name -- name of wavelet filter (character string). * Class -- class of wavelet filters (character string). * Transform -- name of transform (character string). Possible values for transform are: DWT, MODWT.
SEE ALSO
[top]
NAME
WMTSA DWT Structures -- structs used in WMTSA DWT toolbox.
DESCRIPTION
MATLAB structures used in WMTSA DWT toolbox. TOOLBOX wmtsa/dwt CATEGORY
[top]
NAME
dwt_scaling_transfer_function -- Calculate transfer function for frequencies f for specified DWT scaling filter and (optionally) jth level. SYNOPSIS G_j = dwt_scaling_transfer_function(f, wtfname, [j])
INPUTS
f - vector of sinsuoidal frequency. wtfname - name of a WMSTA-supported DWT scaling filter. j - (optional) level (index) of scale.
OUTPUTS
G - vector of the transfer function values for DWT scaling filter h at frequencies f. G_j - vector of the transfer function values for DWT scaling filter h at frequencies f at jth level.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
ALGORITHM
See page 76 of WMTSA for G. See page 97 of WMTSA for G_j.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2003-10-08
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
modwt_scaling_sgf -- Calculate squared gain function for frequencies f for specified DWT scaling filter and (optionally) jth level. SYNOPSIS Gs_j = dwt_scaling_sgf(f, wtfname, [j])
INPUTS
f = vector of sinsuoidal frequency. wtfname = name of a WMSTA-supported DWT scaling filter. j = (optional) level (index) of scale.
OUTPUTS
Gs = vector of squared gain function values for DWT scaling filter h at frequencies f. Gs_j = vector of squared gain function values for DWT scaling filter h at frequencies f at jth level.
SIDE EFFECTS
DESCRIPTION
EXAMPLE
NOTES
ALGORITHM
See page 76 of WMTSA for Gs. See page 154 of WMTSA for Gs_j.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
dwt_scaling_transfer_function
AUTHOR
Charlie Cornish
CREATION DATE
2003-10-08
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
dwt_functionality_tcase -- munit test case to test dwt functionality.
USAGE
run_tcase('dwt_functionality_tcase')
INPUTS
OUTPUTS
tc = tcase structure for dwt testcase.
DESCRIPTION
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-25
COPYRIGHT
(c) 2004, 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
dwt_filter_tcase -- munit test case to test dwt_filter.
USAGE
run_tcase('dwt_filter_tcase')
INPUTS
OUTPUTS
tc = tcase structure for dwt_filter testcase.
SIDE EFFECTS
DESCRIPTION
SEE ALSO
dwt_filter
AUTHOR
Charlie Cornish
CREATION DATE
2005-03-01
COPYRIGHT
(c) 2004, 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 615 $
[top]
NAME
dwt_filter -- Define the DWT filter coefficients. SYNOPSIS [wtf] = dwt_filter(wtfname)
INPUTS
* wtfname -- name of wavelet transform filter (string, case-insenstive).
OUTPUTS
* wtf -- wavelet tranform filter struct (wtf_s).
SIDE EFFECTS
wtfname is a valid wavelet filter name; otherwise error.
DESCRIPTION
dwt_filter returns a wtf_s struct containing the DWT wavelet (high-pass) and scaling (low-pass) filter coefficients.
NOTES
dwt_filter is deprecated by the wtfilter function. dwt_filter is a pass thru to wtfilter and maintained for backward compatiablity and convenice. The wtf_s struct has fields: * g -- scaling (low-pass) filter coefficients (vector). * h -- wavelet (high-pass) filter coefficients (vector). * L -- filter length (= number of coefficients) (integer). * name -- name of wavelet filter (string). * wtfclass -- class of wavelet filters (string). * transform -- name of transform (string). Typing dwt_filter('list') displays a list of supported filters. Typing dwt_filter('all') returns a struct array of wtf_s of all supported filters.
ERRORS
WMTSA:InvalidNumArguments
EXAMPLE
[h, g, L, name] = dwt_filter('la8');
NOTES
dwt_filter is a wrapper function around the wtfilter function.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
wtfilter, modwt_filter TOOLBOX wmtsa/dwt CATEGORY Filters: Filters
AUTHOR
Charlie Cornish Brandon Whitcher
CREATION DATE
2003-09-18
COPYRIGHT
(c) Charles R. Cornish 2005
CREDITS
Based on the original function (myfilter.m) by Brandon Whitcher.
REVISION
$Revision: 612 $
[top]
NAME
dwt_equivalent_filter -- Calculate DWT equivalent filter for J levels. SYNOPSIS [hJ, gJ, LJ] = dwt_equivalent_filter(wtfname, J)
INPUTS
wtfname = string containing name of a WMTSA-supported wavelet filter, case-insensitve. J = number of levels (integer > 0)
OUTPUTS
hJ = 1xJ numeric cell array containing equivalent wavelet filter coefficients. gJ = 1xJ numeric cell array containing equivalent scaling filter coefficients. LJ = 1xJ numeric cell array containing widths of the equivalent filters (L_j).
SIDE EFFECTS
DESCRIPTION
EXAMPLE
WARNINGS
ERRORS
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
2004-01-27
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
dwt2vector -- Convert DWT coefficient cell array to vector representation. SYNOPSIS [W] = dwt2vector(WJ, VJ, att)
INPUTS
* WJ -- DWT wavelet coefficents (cell array of length J0 of NJ(:,1) x NChan matrices). * VJ -- DWT scaling coefficents (cell array of length J0 of NJ(:,2) x NChan matrices). * att -- structure containing DWT transform attributes.
OUTPUTS
* W -- vector of DWT coefficients. (vector or cell array of vectors).
SIDE EFFECTS
DESCRIPTION
USAGE
WARNINGS
ERRORS
EXAMPLE
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
(c) 2005 Charles R. Cornish MATLAB VERSION 7.0
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
dwt -- Compute the (partial) discrete wavelet transform (DWT).
USAGE
[W, V, att, NJ] = dwt(X, [wtf], [nlevels], [boundary], [{opts}])
INPUTS
* X -- set of observations (vector of length N or matrix of size N x Nchan) * wtf -- (optional) wavelet transform filter name or struct (string, case-insensitve or wtf struct). Default: 'la8' * nlevels -- (optional) maximum level J0 (integer) or method of calculating J0 (character string). Valid values: integer>0 or a valid method name Default: 'conservative' * boundary -- (optional) boundary conditions to use (character string) Valid values: 'circular' or 'reflection' Default: 'reflection' * opts -- (optional) Additional function options.
OUTPUTS
* WJ -- DWT wavelet coefficents (cell array of length J0 of NJ(:,1) x NChan matrices). * VJ -- DWT scaling coefficents (cell array of length J0 of NJ(:,2) x NChan matrices). * att -- structure containing DWT transform attributes. * NJ -- Jx2 matrix containing number of DWT wavelet (WJ) and scaling (VJ) coefficients at each level j.
SIDE EFFECTS
1. wavelet is a WMTSA-supported DWT wavelet filter; otherwise error.
DESCRIPTION
EXAMPLE
WARNINGS
ERRORS
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
CREDITS
REVISION
$Revision: 630 $
[top]
NAME
dhm -- Generate simulation of stationary process using Davies-Harte method. SYNOPSIS Y_t = wmtsa_gen_stationary_process(s_X, Z)
INPUTS
* s_X -- ACVS of SDF of process (M lags, D deltas). * Z -- N sets M IID Gaussian RVs.
OUTPUTS
* Y_t -- simulated time series.
SIDE EFFECTS
DESCRIPTION
USAGE
WARNINGS
ERRORS
EXAMPLE
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
data_tsuite -- munit test suite to test data.
USAGE
INPUTS
OUTPUTS
ts = tsuite structure for data testsuite.
SIDE EFFECTS
DESCRIPTION
AUTHOR
Charlie Cornish
CREATION DATE
2005-08-01
COPYRIGHT
(c) 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
compare_struct_fieldvalues -- Compare two structs for match of fieldvalues. SYNOPSIS [tf, msg] = compare_struct_fieldvalues(s1, s2) [tf, msg] = compare_struct_fieldvalues(s1, s2, 'exact') [tf, msg] = compare_struct_fieldvalues(s1, s2, 'sorted')
INPUTS
* s1 -- first struct to compare. * s2 -- second struct to compare. * match -- type of match. Possible values: 'exact', 'sorted'
OUTPUTS
* tf -- flag indicating whether structs' fieldvalues match (Boolean). * mismatches -- list of fieldnames with mismatched values.
SIDE EFFECTS
Function call requires a minimum of 2 input arguments; otherwise error.
DESCRIPTION
compare_struct_filenames compares the fieldvalues in two structs.
USAGE
WARNINGS
ERRORS
EXAMPLE
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
(c) 2005 Charles R. Cornish MATLAB VERSION 7.0
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
compare_struct_fieldnames -- Compare two structs for match of fieldnames. SYNOPSIS [tf, msg] = compare_struct_fieldnames(s1, s2) [tf, msg] = compare_struct_fieldnames(s1, s2, 'exact') [tf, msg] = compare_struct_fieldnames(s1, s2, 'sorted')
INPUTS
* s1 -- first struct to compare. * s2 -- second struct to compare. * match -- type of match. Possible values: 'exact', 'sorted'
OUTPUTS
* tf -- flag indicating whether structs' fieldnames match (Boolean). * mismatches -- list of pairs of non-matching fieldnames (Nx2 cell array of strings).
SIDE EFFECTS
Function call requires a minimum of 2 input arguments; otherwise error.
DESCRIPTION
compare_struct_filenames compares the fieldnames in two structs.
USAGE
WARNINGS
ERRORS
EXAMPLE
NOTES
BUGS
TODO
ALGORITHM
REFERENCES
SEE ALSO
TOOLBOX CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
COPYRIGHT
(c) 2005 Charles R. Cornish MATLAB VERSION 7.0
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
columnize -- Convert array in to a column vector. SYNOPSIS y = columnize(X)
INPUTS
* X -- an array.
OUTPUTS
* y -- column vector
DESCRIPTION
columnize converts an array into a column vector. TOOLBOX utils CATEGORY
AUTHOR
Charlie Cornish
CREATION DATE
2005-Sep-11
COPYRIGHT
(c) 2005 Charles R. Cornish MATLAB VERSION 7.0
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
colorspecname2rgb -- Look up RGB value by ColorSpec name.
USAGE
[rgb] = colorspecname2rgb(colorspecname)
INPUTS
colorspecname - name of ColorSpec in specified format, in short or long format.
OUTPUTS
rbg - three-element row vector whose elements specify the intensities of the red, green, and blue components of the color; the intensities must be in the range [0 1].
SIDE EFFECTS
rgb must be RGB value for one of eight primary colors; otherwise error.
DESCRIPTION
Funciton converts a ColorSpec name, in either short (single character) or long (single word) format into RGB 3-element vector to specifying its RGB value. Possible ColorSpec names: yellow magenta cyan red green blue white black
SEE ALSO
ColorSpec TOOLBOX wmtsa CATEGORY utils
AUTHOR
Charlie Cornish
CREATION DATE
2004-Apr-05
COPYRIGHT
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
argterr_tcase -- munit test case to test argterr.
USAGE
run_tcase(@argterr_tcase)
INPUTS
OUTPUTS
tc = tcase structure for argterr testcase.
SIDE EFFECTS
DESCRIPTION
AUTHOR
Charlie Cornish
CREATION DATE
2005-07-18
COPYRIGHT
(c) 2005 Charles R. Cornish
CREDITS
REVISION
$Revision: 612 $
[top]
NAME
argterr - Check the data type and size of a function argument. SYNOPSIS [errmsg] = argterr(func, arg, type, [arg_size], [mode]) [errmsg] = argterr(func, arg, type, [arg_size], [mode], [msg], 'string') [errstruct] = argterr(func, arg, type, [arg_size], [mode], [msg], 'struct')
INPUTS
* func -- checked function (string or function handle). * arg -- the argument to check (object) * datatypes -- expected data type(s) of the arg. See verify_datatypes function for possibles datatypes to check. (string or string cell array). * arg_size -- (optional) expected size of array (integer vector). * mode -- (optional) output display mode (charcter string) * msg -- (optional) message string to be displayed (string). * return_type -- (optional) type of output argument (string). 'string' = return error message string (default). 'struct' = return error message struct.
OUTPUTS
* errmsg -- error message (string). * errstruct -- error struct with fields: message, identifier.
SIDE EFFECTS
Function call requires a minimum of three input arguments; otherwise error.
DESCRIPTION
argterr checks the data type(s) and optionally the size of an argument to a function call. If arg does not have the specified data type(s), an error message or error struct is returned. arg may be a single Possible values for 'datatypes' to check include: * 'posint' -- All are positive integers --> integer value(s) > 0. * 'int0' -- All are positive integers plus zero --> integer value(s) >= 0. * 'int','integer' -- All are integers --> any integer value(s). * 'num','numeric' -- All are numeric. * 'struct','structure' -- Is a structure. * 'char','character','string' - Is a character string. * 'scalar -- Is a point (size of all dimensions = 1). * 'vec','vector' -- Is a vector (i.e. MxN, with M and/or N = 1). * 'nonsingleton','truevector' -- Is a vector (i.e. MxN with M *or* N = 1). * 'row','rowvector' -- Is a row vector (i.e. M x 1). * 'col','columnvector' -- Is a column vector (i.e. 1 x N). * 'finite' -- All are finite. * 'nonsparse' -- Is a non-sparse matrix. The input argument 'mode' controls the display of diagnostic information to the consolue and has possible values: * 0 -- silent * 1 -- verbose The default value is 1 (silent).
EXAMPLE
arg = [0 2]; % arg is an integer. err = argterr('myfunction', arg, 'posint') % Result: err = 1 err = argterr('myfunction', arg, 'int', [1 2], '') % Result: err = 0 [err, errmsg] = argterr('myfunction', arg, {'int', 'vector'}) % Result: err = 0
NOTES
argterr is modelled after errargt, which is part of MATLAB wavelet toolkit.
SEE ALSO
verify_datatype
AUTHOR
Charlie Cornish
CREATION DATE
2003-06-22
COPYRIGHT
(c) 2003, 2004, 2005 Charles R. Cornish
CREDITS
argterr is inspired by the function errargt by M. Misiti, Y. Misiti, G. Oppenheim, J.M. Poggi which is part of the MATLAB wavelet toolkit. MATLAB VERSION 7.0
REVISION
$Revision: 612 $
[top]
NAME
advance_wavelet_filter -- Calculate the advance of the wavelet filter at jth level for a given wavelet. SYNOPSIS nuHj = advance_wavelet_filter(wtfname, j)
INPUTS
wtfname = string containing name of WMTSA-supported wavelet filter. j = jth level (index) of scale or a range of j levels of scales (integer or Jx1 vector of integers).
OUTPUTS
nuHj = advance of wavelet filter at jth level (integer or vector of integers).
SIDE EFFECTS
wavelet is a WMTSA-supported wavelet filter; otherwise error.
DESCRIPTION
EXAMPLE
ALGORITHM
nuHj = - (2^(j-1) * (L-1) + nu); For details, see equation 114b of WMTSA.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
advance_time_series_filter, dwt_filter
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-08
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
advance_time_series_filter -- Calculate the advance of the time series or filter for a given wavelet. SYNOPSIS nu = advance_time_series_filter(wtfname)
INPUTS
wtfname - string containing name of WMTSA-supported wavelet filter.
OUTPUTS
nu - advance of time series for specified wavelet filter.
SIDE EFFECTS
wavelet is a WMTSA-supported wavelet filter; otherwise error.
DESCRIPTION
EXAMPLE
ALGORITHM
For Least Asymmetric filters, equation 112e of WMTSA: nu = -L/2 + 1, for L/2 is even; = -L/2, for L = 10 or 18; = -L/2 + 2, for L = 14. For Best Localized filter, page 119 of WMTSA. nu = -5, for L = 14; = -11, for L = 18; = -9, for L = 20. For Coiflet filters, page 124 and equation 124 of WMTSA: nu = -2*L/3 + 1
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
dwt_filter
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-08
COPYRIGHT
REVISION
$Revision: 612 $
[top]
NAME
advance_scaling_filter -- Calculate the value to advance scaling filter at jth level for a given wavelet. SYNOPSIS nuGj = advance_scaling_filter(wtfname, level)
INPUTS
wtfname = string containing name of WMTSA-supported wavelet filter. j = jth level (index) of scale or a range of j levels of scales (integer or vector of integers).
OUTPUTS
nuGj = advance of scaling filter at specified levels.
SIDE EFFECTS
wavelet is a WMTSA-supported scaling filter; otherwise error.
DESCRIPTION
EXAMPLE
ALGORITHM
nuGj = (2^j - 1) * nu For details, see equation 114a of WMTSA.
REFERENCES
Percival, D. B. and A. T. Walden (2000) Wavelet Methods for Time Series Analysis. Cambridge: Cambridge University Press.
SEE ALSO
advance_time_series_filter, dwt_filter
AUTHOR
Charlie Cornish
CREATION DATE
2003-05-08
COPYRIGHT
REVISION
$Revision: 612 $