genfis1

(To be removed) Generate fuzzy inference system from data using grid partition

genfis1 will be removed in a future release. Use genfis instead. For more information, see Compatibility Considerations.

Syntax

fis = genfis1(data)

fis = genfis1(data,numMFs)

fis = genfis1(data,numMFs,inmftype)

fis = genfis1(data,numMFs,inmftype,outmftype)

Description

fis = genfis1(data) returns a single-output Sugeno fuzzy inference system (FIS) using a grid partition of the given training data.

fis = genfis1(data,numMFs) specifies the number of membership functions to use for each input variable.

fis = genfis1(data,numMFs,inmftype) specifies the type of membership function to use for input variables.

fis = genfis1(data,numMFs,inmftype,outmftype) specifies the type of membership function to use for output variables.

Examples

Generate FIS Using Grid Partitioning

Generate a FIS using grid partitioning.

data = [rand(10,1) 10*rand(10,1)-5 rand(10,1)];
numMFs = [3 7];
mfType = char('pimf','trimf');
fis = genfis1(data,numMFs,mfType);

To see the contents of fis, use showfis(fis).

Plot the FIS input membership functions.

[x,mf] = plotmf(fis,'input',1);
subplot(2,1,1)
plot(x,mf)
xlabel('input 1 (pimf)')
[x,mf] = plotmf(fis,'input',2);
subplot(2,1,2)
plot(x,mf)
xlabel('input 2 (trimf)')

Input Arguments

data — Training data

array

Training data, specified as an array with M + 1 columns. The first M columns represent input variable values and the last column represents the output variable value. The number of rows is the number of training data points.

numMFs — Number of membership functions for input variables

2 (default) | vector of positive integers | positive integer

Number of membership functions for input variables, specified as a vector of positive integers. Each element of numMFs corresponds to a given input variable. To use the same number of membership functions for all input variables, specify numMFs as a positive integer.

inmftype — Input variable membership function type

'gbellmf' (default) | string | character vector | character array

Input variable membership function type, specified as a string or character vector to use the same membership function for all input variables, or as a character array to use different membership functions for each input variable.

When you specify a character array, each row specifies the membership function type for one input variable.

The following table lists possible membership functions.

Membership function type	Description	For more information
'gbellmf'	Generalized bell-shaped membership function	gbellmf
'gaussmf'	Gaussian membership function	gaussmf
'gauss2mf'	Gaussian combination membership function	gauss2mf
'trimf'	Triangular membership function	trimf
'trapmf'	Trapezoidal membership function	trapmf
'sigmf'	Sigmoidal membership function	sigmf
'dsigmf'	Difference between two sigmoidal membership functions	dsigmf
'psigmf'	Product of two sigmoidal membership functions	psigmf
'zmf'	Z-shaped membership function	zmf
'pimf'	Pi-shaped membership function	pimf
'smf'	S-shaped membership function	smf

outmftype — Output variable membership function type

'linear' (default) | 'constant'

Output variable membership function type, specified as either 'linear' or 'constant'. The number of membership functions associated with the output is the same as the number of rules generated by genfis1.

Output Arguments

fis — Fuzzy inference system

sugfis object

Fuzzy inference system, returned as a sugfis object.

The following table shows the default inference methods for this fuzzy system.

Inference Method	Default
AND	prod
OR	max
Implication	prod
Aggregation	max
Defuzzification	wtaver

Version History

Introduced before R2006a

R2019b: Support for fuzzy inference system structures will be removed

Support for representing fuzzy inference systems as structures will be removed in a future release. Use mamfis and sugfis objects with this function instead. To convert existing fuzzy inference system structures to objects, use the convertfis function.

This change was announced in R2018b. Using fuzzy inference system structures with this function issues a warning starting in R2019b.

R2017a: To be removed

genfis1 will be removed in a future release. Use genfis instead. There are differences between these functions that require updates to your code.

Update Code

To generate a fuzzy system using grid partitioning, first create a default genfisOptions set.

opt = genfisOptions('GridPartition');

You can modify the options using dot notation. Any options you do not modify remain at their default values.

Then, update your code to use genfis. For example, suppose your code has the following form.

fis = genfis1(data,numMFs,inmftype,outmftype);

Use the following code instead.

opt = genfisOptions('GridPartition');
opt.NumMembershipFunctions = numMFs;
opt.InputMembershipFunctionType = inmftype;
opt.OutputMembershipFunctionType = outmftype;
inputData = data(:,end-1);
outputData = data(:,end);
fis = genfis(inputData,outputData,opt);