Comments and Ratings

Apparently, the two-star ranking stems from Alessandro not finding the XLS file. Does anybody else not see it? Please give me a shout.

The file SAS-Matlab Converter.xls is not available for download

Hello,
Great work...
Just a question: Do you know what scilab function ( used with wich parameters) could replace your fminbnd in order to get the fit.
I just can t get the propper algorithm on scilab...

Thank you! Very helpful.
If it can do winsorize by group that would be better!

Just what I was looking for. Wonder why MW released a 'join' (kinda) for datasets but not for matrices.

Not sure what happened; seems to work fine on my side.

.. The file shows up as empty; I am going to re-submit it now.

Hello. Thanks for your code, is very usefull, because I have the need to create synthetic data for work with optimzation. Have you ever been working in the optimization?

I am lucky, I guess. Worked ok, except the warning that Dimitri mentions.

Keep up the good work, non-engineer :)

Hi,
I got an error when pressing Go, after having generated the time series :
"
??? Index exceeds matrix dimensions.

Error in ==> hedgedemo>changeHedgeAssetSeries/changeHedgeValueSeries at 411
        tradeValue = tradeQuant.*aPrice(3:4,:);

Error in ==> hedgedemo>changeHedgeAssetSeries at 370
    changeHedgeValueSeries

Error in ==> hedgedemo>fillFrame41/addFreqPanel/constructHedge at 184
            changeHedgeAssetSeries

??? Error while evaluating uicontrol Callback
"

and sometimes
"
Warning: Matrix is singular, close to singular or badly scaled.
         Results may be inaccurate. RCOND = NaN.
> In hedgedemo>changeHedgeAssetSeries/changeHedgeQuantSeries at 379
  In hedgedemo>changeHedgeAssetSeries at 369
  In hedgedemo>fillFrame41/addFreqPanel/constructHedge at 184
??? Index exceeds matrix dimensions.
"

It does that, even when using inputs displayed on the jpg example.
Do you have an idea ?

Note NELSONFIT's default (and easy-to-change) behavior of searching for tau in (0,10) and adjust the range if measuring time in days or months. (The Nelson-Sielgel function is homogeneous of degree 0 in x/tau).

altogether a very nice, professional demo script showing various ways to visualize a sparse matrix including some nice tricks for the newcomer on how to create spiffy figures and a prototypical macro for a quick axis setup.
unfortunately, the author HAS to use a construct that includes a call to EVAL, which is solely due to the fact that TMW (after MANY years of asking for) has not yet had the time to implement the SET(handle,...) command with a return argument in all cases... otherwise, the community could finally use a function handle of this kind:

fh=@(varargin) {
set(gca,'xtick',varargin{1}) % NOT POSSIBLE!
xlabel(varargin{2}) % OK
title(varargin{3}) % OK
};

us

just a thought:

the eval(touchup) is a bit cumbersome (especially in view of your well documented ML skills...); why not use a construct like this

% only two prop/val are shown
% for brevity...
tf=@() set(gca,'xtick',1:k,'xlim',xlim);
% and use it
tf();

us

This is pretty neat Dimitri. Specially for teaching (that's probably will you did it).

But I have a question.
From the output, it seems to me that GA is performing worst than direct optimization, is that right ?
This sound intuitive because port optimization is a pretty simple mathematical problem and it seems to me that GA doesn't really fit the problem.

Anyway, just with direct opt, still a great submission.

Another thing, please delete the 'close all' in the end. I wanted to see the plots, not close them.

And yes, I also found the picture pretty cool :)

.. In the event, I decided to spin off MBS stuff (albeit not including an IO/PO example) into a separate submission. Please see CFEVALDEMO2 at http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objectId=16570&objectType=FILE

Erratum: 'sasfile and xlsfile must provide absolute file paths; file extensions may NOT be omitted'.

Actually, yes; probably some very simple IO/PO valuation stuff.

A simple tool, but it has everything that any good code should. Good help, an example, an H1 line, complete error checks, it is vectorized, etc.

Link to BSTRAP: http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objectId=14664&objectType=FILE

Bug fix: line 122 should read
res = nan(n1,n2,b+1);
('n' replaced with 'b')

Thanks a lot!