You can subscribe to this list here.
2004 
_{Jan}

_{Feb}

_{Mar}

_{Apr}

_{May}

_{Jun}

_{Jul}

_{Aug}
(1) 
_{Sep}

_{Oct}

_{Nov}
(1) 
_{Dec}


2005 
_{Jan}

_{Feb}

_{Mar}

_{Apr}

_{May}

_{Jun}
(1) 
_{Jul}

_{Aug}

_{Sep}

_{Oct}
(2) 
_{Nov}

_{Dec}
(1) 
2006 
_{Jan}

_{Feb}

_{Mar}

_{Apr}

_{May}
(3) 
_{Jun}
(1) 
_{Jul}
(3) 
_{Aug}
(8) 
_{Sep}

_{Oct}

_{Nov}

_{Dec}

2007 
_{Jan}
(1) 
_{Feb}

_{Mar}
(1) 
_{Apr}

_{May}
(2) 
_{Jun}
(3) 
_{Jul}
(1) 
_{Aug}
(4) 
_{Sep}
(15) 
_{Oct}
(4) 
_{Nov}

_{Dec}

2008 
_{Jan}
(10) 
_{Feb}
(2) 
_{Mar}

_{Apr}

_{May}
(7) 
_{Jun}
(4) 
_{Jul}
(6) 
_{Aug}
(12) 
_{Sep}

_{Oct}
(3) 
_{Nov}
(13) 
_{Dec}
(10) 
2009 
_{Jan}
(12) 
_{Feb}
(19) 
_{Mar}
(27) 
_{Apr}

_{May}
(6) 
_{Jun}
(9) 
_{Jul}

_{Aug}
(5) 
_{Sep}
(12) 
_{Oct}
(20) 
_{Nov}
(1) 
_{Dec}
(8) 
2010 
_{Jan}
(5) 
_{Feb}
(8) 
_{Mar}
(3) 
_{Apr}
(4) 
_{May}
(3) 
_{Jun}
(12) 
_{Jul}
(22) 
_{Aug}
(19) 
_{Sep}
(7) 
_{Oct}
(7) 
_{Nov}
(7) 
_{Dec}
(21) 
2011 
_{Jan}
(10) 
_{Feb}
(18) 
_{Mar}
(26) 
_{Apr}
(12) 
_{May}

_{Jun}
(3) 
_{Jul}
(6) 
_{Aug}
(11) 
_{Sep}
(19) 
_{Oct}
(32) 
_{Nov}
(31) 
_{Dec}
(27) 
2012 
_{Jan}
(8) 
_{Feb}
(5) 
_{Mar}
(19) 
_{Apr}
(3) 
_{May}
(3) 
_{Jun}
(14) 
_{Jul}
(15) 
_{Aug}
(3) 
_{Sep}
(14) 
_{Oct}
(7) 
_{Nov}
(6) 
_{Dec}
(36) 
2013 
_{Jan}
(18) 
_{Feb}
(8) 
_{Mar}
(22) 
_{Apr}
(4) 
_{May}
(18) 
_{Jun}
(16) 
_{Jul}
(9) 
_{Aug}
(8) 
_{Sep}
(4) 
_{Oct}
(6) 
_{Nov}
(1) 
_{Dec}
(3) 
2014 
_{Jan}
(5) 
_{Feb}
(3) 
_{Mar}
(5) 
_{Apr}
(6) 
_{May}
(2) 
_{Jun}

_{Jul}
(4) 
_{Aug}
(4) 
_{Sep}
(7) 
_{Oct}
(3) 
_{Nov}
(5) 
_{Dec}
(3) 
2015 
_{Jan}
(1) 
_{Feb}

_{Mar}

_{Apr}
(1) 
_{May}
(2) 
_{Jun}

_{Jul}
(3) 
_{Aug}
(4) 
_{Sep}

_{Oct}
(5) 
_{Nov}

_{Dec}

2016 
_{Jan}
(11) 
_{Feb}
(7) 
_{Mar}
(1) 
_{Apr}
(1) 
_{May}
(4) 
_{Jun}
(3) 
_{Jul}
(6) 
_{Aug}
(1) 
_{Sep}
(1) 
_{Oct}
(1) 
_{Nov}
(5) 
_{Dec}

2017 
_{Jan}
(6) 
_{Feb}

_{Mar}

_{Apr}

_{May}

_{Jun}

_{Jul}

_{Aug}

_{Sep}

_{Oct}

_{Nov}

_{Dec}

S  M  T  W  T  F  S 






1

2

3

4

5

6

7

8

9

10
(1) 
11
(2) 
12
(2) 
13

14

15

16

17

18
(1) 
19
(1) 
20

21

22

23

24

25

26

27

28

29

30

31







From: Niko Wilbert <mail@ni...>  20101019 11:35:13

Hi everyone, today Sven stumbled over this old issue, and I noticed that we still haven't resolved this. But it seems that we had more or less consensus:  New include_last_sample keyword argument, False for the current behavior, True for including the last point in the covariance.  Set the default to True in the MDP 3.0 release. Cheers, Niko On Fri, Aug 27, 2010 at 4:01 PM, Tiziano Zito <tiziano.zito@...> wrote: >> The advantage of this is that the behavior is easily understandable, >> and all of the scenarios that have been mentioned are still possible. >> It also avoid overcomplicating the node (keeping the last point of >> the last batch in memory, deciding which one is the last batch and >> what to do with the last point of the last batch). > I agree that trying to be too smart would be a problem here. keeping > the last sample in memory would mean that you need for every train > sweep to create a new array with the last sample pasted on top > (arrays can not be appended or inserted onto other arrays without > copying memory around). adding the argument to the train method > seems also a little overkill. the argument to the constructor could > set a public attribute, that the user can overwrite if needed. > we need to come up with a proper name for the argument  and with > proper documentation. I can not come up with a better name than > "include_last_sample" both for the argument and for the attribute. > I'm somehow tending to prefer changing the default to include the > last point. after all with the mdp 3 release we are going to break > backward compatibility for other things too... > what do you think? > > tiziano > > > >  > Sell apps to millions through the Intel(R) Atom(Tm) Developer Program > Be part of this innovative community and reach millions of netbook users > worldwide. Take advantage of special opportunities to increase revenue and > speed timetomarket. Join now, and jumpstart your future. > http://p.sf.net/sfu/intelatomd2d > _______________________________________________ > mdptoolkitusers mailing list > mdptoolkitusers@... > https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers > 
From: Pietro Berkes <berkes@ga...>  20101018 13:12:49

Hi Jason! the attached paper came to my mind, where the authors talk about three algorithms for producing decorrelated components of images: PCA, ZCA and ICA. Basically, once you select a subspace of principal components, within that space you can choose any orthonormal basis and still get decorrelated components that explain the same amount of variance in the input, so the final solution is only determined up to a rotation. PCA is the solution with no rotations, ZCA and ICA choose different rotations (see paper). Is it possible that ZCA is what minitab is doing? If not, I'm out of suggestions... I'm pretty sure MDP is doing the right PCA, though :) Otherwise, is the minitab code accessible and can you post it? That way we can have a look at it instead of guessing. P. On Tue, Oct 12, 2010 at 12:35 PM, Jason Merrill <jason.merrill@...> wrote: > Hi Pietro, > > Thanks for the tips. I had a look at WhiteningNode, and I find that > > zscores_whitening = zscores_pca/numpy.sqrt(pca_node.d[0]) > whitening_node.v = pca_node.v/numpy.sqrt(pca_node.d[0]) > > whitening_node.v = whitening_node.get_projmatrix() > > That is, the WhiteningNode just rescales everything from PCANode by > the standard deviation, as I had done manually. > > The projection vector from MDP is not parallel to the PCA coefficients > from minitab. I think I need to investigate exactly what the meaning > of the Loadings and Coefficients are in minitab. > > I've added a column to the spreadsheet showing zscores_whitening = > zscores_pca/numpy.sqrt(pca_node.d[0]) > > https://spreadsheets.google.com/ccc?key=0AjmRxgfbN9spdHZockhyTHpRclJndXFqekRBbTJQX3c&hl=en&authkey=CIP7_awI#gid=0 > > Regards, > > JM > > On Tue, Oct 12, 2010 at 7:30 AM, Pietro Berkes <berkes@...> wrote: >> Hi Jason, >> >> I don't have any experience with minitab, but here are a couple of >> things to check: >> >> 1) a better way to compare the two algorithm is probably to check that >> the projection matrices are the same; in mdp, you find the projection >> matrix as >> a = pca_node.get_projection_matrix(transposed=False) >> then if you do >> numpy.dot(a.T, proj_matrix_from_minitab) >> you should obtain a matrix with only one element per row active (if >> the directions of projections are the same) and that element should be >> 1 if they also have the same length >> >> 2) make sure that minitab also sorts the principal components from >> largest to smallest eigenvalue >> >> 3) I'm not sure at the moment how Zscores PCA work, but my guess is >> that it is equivalent to the WhiteningNode rather than to the PCANode >> (WhiteningNode rescales the components such that the output variance >> is one). Could you please check the output of WhiteningNode? >> >> Let me know how it goes, I'll dig deeper if none of the above helps. >> Best, >> Pietro >> >> >> On Mon, Oct 11, 2010 at 9:24 AM, Jason Merrill <jason.merrill@...> wrote: >>>> just a quick guess, but the PCA node has an svd keyword which should >>>> make it numerically more stable I believe, did you try using it with >>>> svd=True? >>> >>> Just gave that a try, but in my case it does not seem to change the >>> answers at all. Or at least not in the first six digits. >>> >>> JM >>> >>>> Regards, >>>> >>>> Sebastian >>>> >>>> >>>> On Sun, 20101010 at 19:24 0400, Jason Merrill wrote: >>>>> I'm trying to write some statistical analysis in MDP that I had been >>>>> doing in Minitab in order to automate it. I'm using factor analysis to >>>>> look at some survey results. In Minitab, there are two options for >>>>> factor analysis: PCA, and maximum likelihood. I had been using PCA. >>>>> The results from FANode() match the Minitab results using maximum >>>>> likelihood to within the tolerances I specified, but I'd like to match >>>>> the PCA factor analysis results from Minitab. I tried using PCANode, >>>>> and I found that the results are in rough agreement, but don't match >>>>> the Minitab results to within acceptable tolerances. >>>>> >>>>> Does anyone have experience with both pieces of software? Is there a >>>>> way to coax MDP into doing an analysis like Minitab's "pca factor >>>>> analysis"? >>>>> >>>>> I've shared some sample data I'm trying to analyze as a google spreadsheet: >>>>> >>>>> https://spreadsheets.google.com/ccc?key=0AjmRxgfbN9spdHZockhyTHpRclJndXFqekRBbTJQX3c&hl=en&authkey=CIP7_awI#gid=0 >>>>> >>>>> Sheet 1 is the data. Sheet 2 compares the results I get from Minitab >>>>> and MDP for PCA analysis and Maximum Likelihood factor analysis. I've >>>>> pasted the python code for the two different analyses: >>>>> >>>>> Maximum Likelihood: http://pastebin.com/RAqScYPW >>>>> PCA: http://pastebin.com/K8Dv0svg >>>>> >>>>> I'd really appreciate your insight. >>>>> >>>>> Regards, >>>>> >>>>> Jason Merrill >>>>> >>>>>  >>>>> Beautiful is writing same markup. Internet Explorer 9 supports >>>>> standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. >>>>> Spend less time writing and rewriting code and more time creating great >>>>> experiences on the web. Be a part of the beta today. >>>>> http://p.sf.net/sfu/beautyoftheweb >>>>> _______________________________________________ >>>>> mdptoolkitusers mailing list >>>>> mdptoolkitusers@... >>>>> https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers >>>>> >>>> >>>> >>>> >>>>  >>>> Beautiful is writing same markup. Internet Explorer 9 supports >>>> standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. >>>> Spend less time writing and rewriting code and more time creating great >>>> experiences on the web. Be a part of the beta today. >>>> http://p.sf.net/sfu/beautyoftheweb >>>> _______________________________________________ >>>> mdptoolkitusers mailing list >>>> mdptoolkitusers@... >>>> https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers >>>> >>> >>>  >>> Beautiful is writing same markup. Internet Explorer 9 supports >>> standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. >>> Spend less time writing and rewriting code and more time creating great >>> experiences on the web. Be a part of the beta today. >>> http://p.sf.net/sfu/beautyoftheweb >>> _______________________________________________ >>> mdptoolkitusers mailing list >>> mdptoolkitusers@... >>> https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers >>> >> >>  >> Beautiful is writing same markup. Internet Explorer 9 supports >> standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. >> Spend less time writing and rewriting code and more time creating great >> experiences on the web. Be a part of the beta today. >> http://p.sf.net/sfu/beautyoftheweb >> _______________________________________________ >> mdptoolkitusers mailing list >> mdptoolkitusers@... >> https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers >> > >  > Beautiful is writing same markup. Internet Explorer 9 supports > standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. > Spend less time writing and rewriting code and more time creating great > experiences on the web. Be a part of the beta today. > http://p.sf.net/sfu/beautyoftheweb > _______________________________________________ > mdptoolkitusers mailing list > mdptoolkitusers@... > https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers > 
From: Jason Merrill <jason.merrill@ya...>  20101012 16:35:42

Hi Pietro, Thanks for the tips. I had a look at WhiteningNode, and I find that zscores_whitening = zscores_pca/numpy.sqrt(pca_node.d[0]) whitening_node.v = pca_node.v/numpy.sqrt(pca_node.d[0]) whitening_node.v = whitening_node.get_projmatrix() That is, the WhiteningNode just rescales everything from PCANode by the standard deviation, as I had done manually. The projection vector from MDP is not parallel to the PCA coefficients from minitab. I think I need to investigate exactly what the meaning of the Loadings and Coefficients are in minitab. I've added a column to the spreadsheet showing zscores_whitening = zscores_pca/numpy.sqrt(pca_node.d[0]) https://spreadsheets.google.com/ccc?key=0AjmRxgfbN9spdHZockhyTHpRclJndXFqekRBbTJQX3c&hl=en&authkey=CIP7_awI#gid=0 Regards, JM On Tue, Oct 12, 2010 at 7:30 AM, Pietro Berkes <berkes@...> wrote: > Hi Jason, > > I don't have any experience with minitab, but here are a couple of > things to check: > > 1) a better way to compare the two algorithm is probably to check that > the projection matrices are the same; in mdp, you find the projection > matrix as > a = pca_node.get_projection_matrix(transposed=False) > then if you do > numpy.dot(a.T, proj_matrix_from_minitab) > you should obtain a matrix with only one element per row active (if > the directions of projections are the same) and that element should be > 1 if they also have the same length > > 2) make sure that minitab also sorts the principal components from > largest to smallest eigenvalue > > 3) I'm not sure at the moment how Zscores PCA work, but my guess is > that it is equivalent to the WhiteningNode rather than to the PCANode > (WhiteningNode rescales the components such that the output variance > is one). Could you please check the output of WhiteningNode? > > Let me know how it goes, I'll dig deeper if none of the above helps. > Best, > Pietro > > > On Mon, Oct 11, 2010 at 9:24 AM, Jason Merrill <jason.merrill@...> wrote: >>> just a quick guess, but the PCA node has an svd keyword which should >>> make it numerically more stable I believe, did you try using it with >>> svd=True? >> >> Just gave that a try, but in my case it does not seem to change the >> answers at all. Or at least not in the first six digits. >> >> JM >> >>> Regards, >>> >>> Sebastian >>> >>> >>> On Sun, 20101010 at 19:24 0400, Jason Merrill wrote: >>>> I'm trying to write some statistical analysis in MDP that I had been >>>> doing in Minitab in order to automate it. I'm using factor analysis to >>>> look at some survey results. In Minitab, there are two options for >>>> factor analysis: PCA, and maximum likelihood. I had been using PCA. >>>> The results from FANode() match the Minitab results using maximum >>>> likelihood to within the tolerances I specified, but I'd like to match >>>> the PCA factor analysis results from Minitab. I tried using PCANode, >>>> and I found that the results are in rough agreement, but don't match >>>> the Minitab results to within acceptable tolerances. >>>> >>>> Does anyone have experience with both pieces of software? Is there a >>>> way to coax MDP into doing an analysis like Minitab's "pca factor >>>> analysis"? >>>> >>>> I've shared some sample data I'm trying to analyze as a google spreadsheet: >>>> >>>> https://spreadsheets.google.com/ccc?key=0AjmRxgfbN9spdHZockhyTHpRclJndXFqekRBbTJQX3c&hl=en&authkey=CIP7_awI#gid=0 >>>> >>>> Sheet 1 is the data. Sheet 2 compares the results I get from Minitab >>>> and MDP for PCA analysis and Maximum Likelihood factor analysis. I've >>>> pasted the python code for the two different analyses: >>>> >>>> Maximum Likelihood: http://pastebin.com/RAqScYPW >>>> PCA: http://pastebin.com/K8Dv0svg >>>> >>>> I'd really appreciate your insight. >>>> >>>> Regards, >>>> >>>> Jason Merrill >>>> >>>>  >>>> Beautiful is writing same markup. Internet Explorer 9 supports >>>> standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. >>>> Spend less time writing and rewriting code and more time creating great >>>> experiences on the web. Be a part of the beta today. >>>> http://p.sf.net/sfu/beautyoftheweb >>>> _______________________________________________ >>>> mdptoolkitusers mailing list >>>> mdptoolkitusers@... >>>> https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers >>>> >>> >>> >>> >>>  >>> Beautiful is writing same markup. Internet Explorer 9 supports >>> standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. >>> Spend less time writing and rewriting code and more time creating great >>> experiences on the web. Be a part of the beta today. >>> http://p.sf.net/sfu/beautyoftheweb >>> _______________________________________________ >>> mdptoolkitusers mailing list >>> mdptoolkitusers@... >>> https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers >>> >> >>  >> Beautiful is writing same markup. Internet Explorer 9 supports >> standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. >> Spend less time writing and rewriting code and more time creating great >> experiences on the web. Be a part of the beta today. >> http://p.sf.net/sfu/beautyoftheweb >> _______________________________________________ >> mdptoolkitusers mailing list >> mdptoolkitusers@... >> https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers >> > >  > Beautiful is writing same markup. Internet Explorer 9 supports > standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. > Spend less time writing and rewriting code and more time creating great > experiences on the web. Be a part of the beta today. > http://p.sf.net/sfu/beautyoftheweb > _______________________________________________ > mdptoolkitusers mailing list > mdptoolkitusers@... > https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers > 
From: Pietro Berkes <berkes@ga...>  20101012 11:30:34

Hi Jason, I don't have any experience with minitab, but here are a couple of things to check: 1) a better way to compare the two algorithm is probably to check that the projection matrices are the same; in mdp, you find the projection matrix as a = pca_node.get_projection_matrix(transposed=False) then if you do numpy.dot(a.T, proj_matrix_from_minitab) you should obtain a matrix with only one element per row active (if the directions of projections are the same) and that element should be 1 if they also have the same length 2) make sure that minitab also sorts the principal components from largest to smallest eigenvalue 3) I'm not sure at the moment how Zscores PCA work, but my guess is that it is equivalent to the WhiteningNode rather than to the PCANode (WhiteningNode rescales the components such that the output variance is one). Could you please check the output of WhiteningNode? Let me know how it goes, I'll dig deeper if none of the above helps. Best, Pietro On Mon, Oct 11, 2010 at 9:24 AM, Jason Merrill <jason.merrill@...> wrote: >> just a quick guess, but the PCA node has an svd keyword which should >> make it numerically more stable I believe, did you try using it with >> svd=True? > > Just gave that a try, but in my case it does not seem to change the > answers at all. Or at least not in the first six digits. > > JM > >> Regards, >> >> Sebastian >> >> >> On Sun, 20101010 at 19:24 0400, Jason Merrill wrote: >>> I'm trying to write some statistical analysis in MDP that I had been >>> doing in Minitab in order to automate it. I'm using factor analysis to >>> look at some survey results. In Minitab, there are two options for >>> factor analysis: PCA, and maximum likelihood. I had been using PCA. >>> The results from FANode() match the Minitab results using maximum >>> likelihood to within the tolerances I specified, but I'd like to match >>> the PCA factor analysis results from Minitab. I tried using PCANode, >>> and I found that the results are in rough agreement, but don't match >>> the Minitab results to within acceptable tolerances. >>> >>> Does anyone have experience with both pieces of software? Is there a >>> way to coax MDP into doing an analysis like Minitab's "pca factor >>> analysis"? >>> >>> I've shared some sample data I'm trying to analyze as a google spreadsheet: >>> >>> https://spreadsheets.google.com/ccc?key=0AjmRxgfbN9spdHZockhyTHpRclJndXFqekRBbTJQX3c&hl=en&authkey=CIP7_awI#gid=0 >>> >>> Sheet 1 is the data. Sheet 2 compares the results I get from Minitab >>> and MDP for PCA analysis and Maximum Likelihood factor analysis. I've >>> pasted the python code for the two different analyses: >>> >>> Maximum Likelihood: http://pastebin.com/RAqScYPW >>> PCA: http://pastebin.com/K8Dv0svg >>> >>> I'd really appreciate your insight. >>> >>> Regards, >>> >>> Jason Merrill >>> >>>  >>> Beautiful is writing same markup. Internet Explorer 9 supports >>> standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. >>> Spend less time writing and rewriting code and more time creating great >>> experiences on the web. Be a part of the beta today. >>> http://p.sf.net/sfu/beautyoftheweb >>> _______________________________________________ >>> mdptoolkitusers mailing list >>> mdptoolkitusers@... >>> https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers >>> >> >> >> >>  >> Beautiful is writing same markup. Internet Explorer 9 supports >> standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. >> Spend less time writing and rewriting code and more time creating great >> experiences on the web. Be a part of the beta today. >> http://p.sf.net/sfu/beautyoftheweb >> _______________________________________________ >> mdptoolkitusers mailing list >> mdptoolkitusers@... >> https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers >> > >  > Beautiful is writing same markup. Internet Explorer 9 supports > standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. > Spend less time writing and rewriting code and more time creating great > experiences on the web. Be a part of the beta today. > http://p.sf.net/sfu/beautyoftheweb > _______________________________________________ > mdptoolkitusers mailing list > mdptoolkitusers@... > https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers > 
From: Jason Merrill <jason.merrill@ya...>  20101011 13:24:25

> just a quick guess, but the PCA node has an svd keyword which should > make it numerically more stable I believe, did you try using it with > svd=True? Just gave that a try, but in my case it does not seem to change the answers at all. Or at least not in the first six digits. JM > Regards, > > Sebastian > > > On Sun, 20101010 at 19:24 0400, Jason Merrill wrote: >> I'm trying to write some statistical analysis in MDP that I had been >> doing in Minitab in order to automate it. I'm using factor analysis to >> look at some survey results. In Minitab, there are two options for >> factor analysis: PCA, and maximum likelihood. I had been using PCA. >> The results from FANode() match the Minitab results using maximum >> likelihood to within the tolerances I specified, but I'd like to match >> the PCA factor analysis results from Minitab. I tried using PCANode, >> and I found that the results are in rough agreement, but don't match >> the Minitab results to within acceptable tolerances. >> >> Does anyone have experience with both pieces of software? Is there a >> way to coax MDP into doing an analysis like Minitab's "pca factor >> analysis"? >> >> I've shared some sample data I'm trying to analyze as a google spreadsheet: >> >> https://spreadsheets.google.com/ccc?key=0AjmRxgfbN9spdHZockhyTHpRclJndXFqekRBbTJQX3c&hl=en&authkey=CIP7_awI#gid=0 >> >> Sheet 1 is the data. Sheet 2 compares the results I get from Minitab >> and MDP for PCA analysis and Maximum Likelihood factor analysis. I've >> pasted the python code for the two different analyses: >> >> Maximum Likelihood: http://pastebin.com/RAqScYPW >> PCA: http://pastebin.com/K8Dv0svg >> >> I'd really appreciate your insight. >> >> Regards, >> >> Jason Merrill >> >>  >> Beautiful is writing same markup. Internet Explorer 9 supports >> standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. >> Spend less time writing and rewriting code and more time creating great >> experiences on the web. Be a part of the beta today. >> http://p.sf.net/sfu/beautyoftheweb >> _______________________________________________ >> mdptoolkitusers mailing list >> mdptoolkitusers@... >> https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers >> > > > >  > Beautiful is writing same markup. Internet Explorer 9 supports > standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. > Spend less time writing and rewriting code and more time creating great > experiences on the web. Be a part of the beta today. > http://p.sf.net/sfu/beautyoftheweb > _______________________________________________ > mdptoolkitusers mailing list > mdptoolkitusers@... > https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers > 
From: Sebastian Berg <sebastian@si...>  20101011 08:45:43

Hi, just a quick guess, but the PCA node has an svd keyword which should make it numerically more stable I believe, did you try using it with svd=True? Regards, Sebastian On Sun, 20101010 at 19:24 0400, Jason Merrill wrote: > I'm trying to write some statistical analysis in MDP that I had been > doing in Minitab in order to automate it. I'm using factor analysis to > look at some survey results. In Minitab, there are two options for > factor analysis: PCA, and maximum likelihood. I had been using PCA. > The results from FANode() match the Minitab results using maximum > likelihood to within the tolerances I specified, but I'd like to match > the PCA factor analysis results from Minitab. I tried using PCANode, > and I found that the results are in rough agreement, but don't match > the Minitab results to within acceptable tolerances. > > Does anyone have experience with both pieces of software? Is there a > way to coax MDP into doing an analysis like Minitab's "pca factor > analysis"? > > I've shared some sample data I'm trying to analyze as a google spreadsheet: > > https://spreadsheets.google.com/ccc?key=0AjmRxgfbN9spdHZockhyTHpRclJndXFqekRBbTJQX3c&hl=en&authkey=CIP7_awI#gid=0 > > Sheet 1 is the data. Sheet 2 compares the results I get from Minitab > and MDP for PCA analysis and Maximum Likelihood factor analysis. I've > pasted the python code for the two different analyses: > > Maximum Likelihood: http://pastebin.com/RAqScYPW > PCA: http://pastebin.com/K8Dv0svg > > I'd really appreciate your insight. > > Regards, > > Jason Merrill > >  > Beautiful is writing same markup. Internet Explorer 9 supports > standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. > Spend less time writing and rewriting code and more time creating great > experiences on the web. Be a part of the beta today. > http://p.sf.net/sfu/beautyoftheweb > _______________________________________________ > mdptoolkitusers mailing list > mdptoolkitusers@... > https://lists.sourceforge.net/lists/listinfo/mdptoolkitusers > 
From: Jason Merrill <jason.merrill@ya...>  20101010 23:24:55

I'm trying to write some statistical analysis in MDP that I had been doing in Minitab in order to automate it. I'm using factor analysis to look at some survey results. In Minitab, there are two options for factor analysis: PCA, and maximum likelihood. I had been using PCA. The results from FANode() match the Minitab results using maximum likelihood to within the tolerances I specified, but I'd like to match the PCA factor analysis results from Minitab. I tried using PCANode, and I found that the results are in rough agreement, but don't match the Minitab results to within acceptable tolerances. Does anyone have experience with both pieces of software? Is there a way to coax MDP into doing an analysis like Minitab's "pca factor analysis"? I've shared some sample data I'm trying to analyze as a google spreadsheet: https://spreadsheets.google.com/ccc?key=0AjmRxgfbN9spdHZockhyTHpRclJndXFqekRBbTJQX3c&hl=en&authkey=CIP7_awI#gid=0 Sheet 1 is the data. Sheet 2 compares the results I get from Minitab and MDP for PCA analysis and Maximum Likelihood factor analysis. I've pasted the python code for the two different analyses: Maximum Likelihood: http://pastebin.com/RAqScYPW PCA: http://pastebin.com/K8Dv0svg I'd really appreciate your insight. Regards, Jason Merrill 