<html><head></head><body><div class="gmail_quote">On November 30, 2016 7:06:53 PM EST, Salim Al-wasity <salim_alwasity@yahoo.com> wrote:<blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">
1. My datasets have similar number of features and every time I ran (hyper=Hyperalignment  ( )), the reference is the first dataset ds_h (0).<div id="yMail_cursorElementTracker_1480550706625"><br /></div><div id="yMail_cursorElementTracker_1480550708187">2. I also did (hyper= Hyperalignment  (ref_ds=0)) for the second scenario and still getting different result tha scenario 1.</div><div id="yMail_cursorElementTracker_1480550803982"><br /></div><div id="yMail_cursorElementTracker_1480550804876">Sincerely </div><div id="yMail_cursorElementTracker_1480550806558">Salim <br /><br /><div id="ymail_android_signature"><a href="https://overview.mail.yahoo.com/mobile/?.src=Android">Sent from Yahoo Mail on Android</a></div> <br /> <blockquote style="margin: 0 0 20px 0;"> <header style="font-family:Roboto, sans-serif; color:#6D00F6;"> <div>On Wed, 30 Nov, 2016 at 11:14 pm, Yaroslav Halchenko</div><div><yoh@onerussian.com> wrote:</div> </header> <div style="padding:
10px 0 0 20px; margin: 10px 0 0 0; border-left: 1px solid #6D00F6;"> <div class="gmail_quote">On November 30, 2016 5:54:36 PM EST, Salim Al-wasity <salim_alwasity@yahoo.com> wrote:<blockquote class="gmail_quote" style="margin:0pt 0pt 0pt 0.8ex;border-left:1px solid rgb(204, 204, 204);padding-left:1ex;">
<div class="yQTDBase yqt3322313142" id="yqt58842"><div style="color:#000;background-color:#fff;font-family:arial, helvetica, sans-serif;font-size:16px;"><div id="yui_3_16_0_ym19_1_1480546394726_5882">Dears<br clear="none" /></div><div class="qtdSeparateBR"><br clear="none" /><br clear="none" /></div><div class="yahoo_quoted" id="yui_3_16_0_ym19_1_1480546394726_5937" style="display:block;"><div id="yui_3_16_0_ym19_1_1480546394726_5936" style="font-family:arial, helvetica, sans-serif;font-size:16px;"><div id="yui_3_16_0_ym19_1_1480546394726_5935" style="font-family:HelveticaNeue, Helvetica Neue, Helvetica, Arial, Lucida Grande, Sans-Serif;font-size:16px;"><div class="y_msg_container" id="yui_3_16_0_ym19_1_1480546394726_5948"><div id="yiv6131933817"><div id="yui_3_16_0_ym19_1_1480546394726_5947"><div id="yui_3_16_0_ym19_1_1480546394726_5946" style="color:#000;background-color:#fff;font-family:arial, helvetica, sans-serif;font-size:16px;"><div dir="ltr"
id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187">I am running a Hyperalignment analysis to investigate the
effect of the reference subject on the common model and between subject classification. I ran two scenarios to double check my results, and in both cases I got different classification accuracies.If I am not wrong, the below scenarios must giving me identical results</div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187">.</div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187">I had 10 subjects whose Hyperalignment data are stored in (ds_h) and task data in (ds_task) which I am classifying</div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187"><br clear="none" /></div><div id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187">Scenario-1-:</div><div id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187"><br clear="none" /></div><div id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187">hyper=Hyperalignment(ref_ds=2)</div><div dir="ltr"
id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187">hypermaps=hyper(ds_h)</div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187">ds_hyper = [ hypmaps[i].forward(d_all) for i, d_all in enumerate(ds_task)]</div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187">..........Then continue with classification</div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187"><br clear="none" /></div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187"><br clear="none" /></div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187">Scenario-2-:</div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187">new_ds_h=[ds_h[2], ds_h[0], ds_h[1], ds_h[3], ds_h[4], ds_h[5], ds_h[6], ds_h[7], ds_h[8], ds_h[9]]</div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187">hyper=Hyperalignment()</div><div dir="ltr"
id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187">hyper.train(new_ds_h)</div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12187">hypermaps=hyper(ds_h)</div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12566">ds_hyper = [ hypmaps[i].forward(d_all) for i, d_all in enumerate(ds_task)]</div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12567">..........Then continue with classification</div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12568"><br clear="none" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12569" /></div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12568"><br clear="none" /></div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12568">Sincerely</div><div dir="ltr" id="yiv6131933817yui_3_16_0_ym19_1_1480544289680_12568">Salim</div></div></div></div><br clear="none" /><br clear="none" /></div>  </div> </div>  </div></div></div><p
style="margin-top:2.5em;margin-bottom:1em;border-bottom:1px solid #000;"></p><pre class="k9mail"></pre><hr /><br clear="none" />Pkg-ExpPsy-PyMVPA mailing list<br clear="none" />Pkg-ExpPsy-PyMVPA@lists.alioth.debian.org<br clear="none" /><a rel="nofollow" shape="rect" target="_blank" href="http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa">http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa</a></blockquote></div><br clear="all" /><a rel="nofollow" shape="rect" target="_blank" href="http://www.pymvpa.org/generated/mvpa2.algorithms.hyperalignment.Hyperalignment.html#mvpa2.algorithms.hyperalignment.Hyperalignment">http://www.pymvpa.org/generated/mvpa2.algorithms.hyperalignment.Hyperalignment.html#mvpa2.algorithms.hyperalignment.Hyperalignment</a><br clear="none" />
<br clear="none" />
ref_ds : int or None, optional<br clear="none" />
<br clear="none" />
Index of a dataset to use as 1st-level common space reference. If None, then the dataset with the maximum number of features is used. Constraints: (value must be in range [0, inf], and value must be convertible to type ‘int’), or value must be None. [Default: None]<br clear="none" />
<br clear="none" />
So utter is not necessarily the first (0th) dataset which is chosen... So you get the same results if your set ref_ds=0<br clear="none" />
<br clear="none" />
You could enable_ca=["chosen_ref_ds"] and then check ca.chosen_ref_ds on which one is chosen it you don't explicitly specify any<br clear="none" />
-- <br clear="none" />
Sent from a phone which beats iPhone. </div> </blockquote></div><p style="margin-top: 2.5em; margin-bottom: 1em; border-bottom: 1px solid #000"></p><pre class="k9mail"><hr /><br />Pkg-ExpPsy-PyMVPA mailing list<br />Pkg-ExpPsy-PyMVPA@lists.alioth.debian.org<br /><a href="http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa">http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa</a></pre></blockquote></div><br clear="all">Could you quantify how much different?<br>
-- <br>
Sent from a phone which beats iPhone.</body></html>