perClass

Index out of bounds in sdsvc

2012-01-31T22:25:13+01:00

Hi,
training sdsvc with default settings on a particular dataset I obtain as follows:

??? Error using ==> ind2code at 8
Index out of bounds.

Error in ==> ..perClass_Demoperclass@sddatasubset.p>subset at 73

Error in ==> ..perClass_Demoperclass@sddatasubsref.p>subsref at 31

Error in ==> ..perClass_Demoperclasssdsvc.p>sdsvc at 210

Error in ==> ..perClass_Demoperclass@sdpplapply_untrained.p>apply_untrained at 150

Error in ==> ..perClass_Demoperclass@sdpplmtimes.p>mtimes at 72

Error in ==> par_classify_BR at 20
    trainedSvc=trainngData*untrainedSvc;

Could you please help with the problem?

Thanks in advance.
Tomasz

bug in fixedcc.m

2012-01-17T14:49:16+01:00

Hi,

Line 204 in fixedcc.m function :
par - w.data{4};

Should it be replaced by ?
par = w.data{4};

Thank you for your help.

Internal error in sdtree

2012-01-13T17:57:08+01:00

Please help figure out what is the problem.

>> sdtree(ds)
??? Error using ==> perclass_mex
Tree training: Internal error

Error in ==>
C:DocumentsNaukaResearchEnvPerClassperClass_Demoperclasssdtree.p>sdtree at
181

I am using:
perClass Demo 3.1.2 (22-Dec-2011), Copyright (C) 2007-2011, PR Sys Design, All rights reserved

Regards.

nusvc Does not work

2011-08-17T08:47:50+01:00

Example 1:
nusvc([],’p’,1,.1)

generates error due to nu parameter being supplied, but documentation says this should work.

Example 2:
nusvc(dataset(randn(10,3), [ones(5,1);zeros(5,1)]), ‘p’, 1);

generates error as well.

This is on PRTools version 4.1.12

Frustrating…

Inverted axis , title, and x-ylabel using sdscatter

2010-10-24T21:24:19+01:00

Hello,

as you can see at both attached images,
while i was trying to check the
sdscatter(ts,p)
sdscatter(ts,pd2)
according to “Chapter 3: Getting Started”
i received inverted plots..
This also happened during executing the example .m files…

If someone knows a way to fix this please reply…

Thanks..

Small inconsistency in the documentation

2011-06-14T13:44:13+01:00

Dear PRSD Team,

Two minor remarks about Section 9.5.1 in the latest documentation:

1) the “t” of “fruit” is missing in the 4th sentence,
2) In the last line just before Section 9.5.2, I noticed that you’re mentioning a non-cancer classe while demonstrating with the fruit dataset. (I hope I understood correctly :-))

Cheers,

Jean-Michel

Randomly selecting only one sample from the whole dataset

2011-04-08T11:58:07+01:00

Hello,

Using randsubset I can select 1 sample per class but not 1 sample out of the whole dataset.

>> D
14873 by 64 sddata, 10 classes: [1911   197  3923  2078   287   286  3475   509   270  1937]
>> randsubset(D,1)
10 by 64 sddata, 10 classes: [1  1  1  1  1  1  1  1  1  1]
>> randsubset(D,1,'all')
??? Error using ==> randsubset at 115
Parameter out of range

Actually, I do not know yet if it is a misuse or a bug...Could somebody help me?

Best regards,

Jean-Michel

Size of index arrays returned by the subset function

2011-01-03T14:54:54+01:00

Hello,

I noticed that the (index array) output arguments in sddata/subset have inconsistent sizes.


[B,REST,IND,IND2]=subset(A,CLASSES)

While IND is a row vector, IND2 is generally a columns vector.
Example:

>> [r,s,Ir,Is] = subset(a.a,'class','apple');
>> a = load('fruit');
>> [r,s,Ir,Is] = subset(a.a,'class','apple');
>> r
'Fruit set' 100 by 2 sddata, class: 'apple'
>> s
'Fruit set' 160 by 2 sddata, 2 classes: 'banana'(100) 'stone'(60) 
>> size(Ir)

ans =

   100     1

>> size(Is)

ans =

     1   160

Is there a reason for that?

Cheers

Error with ASAM

2010-12-24T12:51:09+01:00

Hi,

Actually I do not know if it is a bug or my mistake.

When training my GMM on my dataset obtained from dataset*sdprox(dataset(proto),’sam’) everything goes smoothly but if I change to ‘asam’ then the learning fails.

Any idea of what it is?

[class 'FA' EM:.............................. 1 comp] [class 'STOP' EM:??? Error using ==> svd
Input to SVD must not contain NaN or Inf.

Error in ==> rank at 15
s = svd(A);

Error in ==> /matlab/PRSD_Studio/prsd_toolbox/private/conv_par.p>conv_par at 294

Error in ==> /matlab/PRSD_Studio/prsd_toolbox/private/conv_ppl.p>process_action at 397

Error in ==> /matlab/PRSD_Studio/prsd_toolbox/private/conv_ppl.p>process at 76

Error in ==> /matlab/PRSD_Studio/prsd_toolbox/private/conv_ppl.p>conv_ppl at 40

Error in ==> /matlab/PRSD_Studio/prsd_toolbox/sdexe.p>sdexe at 70

Error in ==> /matlab/PRSD_Studio/prsd_toolbox/sdmixture.p>em_estimation at 294

Error in ==> /matlab/PRSD_Studio/prsd_toolbox/sdmixture.p>sdmixture at 207

Best regards

Jean-Michel

Licence error

2010-11-20T16:28:50+01:00

Sir,

Am getting the following error.

>> sdscatter(a)

Error checking out license: License has expired (-3)
??? Error using ==> prsd_mex
PRSD License Error

Error in ==> D:PARKsoftPRSD_Studio_Lite_2.0.5_23Dec09_M74PRToolssdscatter.p>sdscatter at 40

Pl. do the needful.