Hi,<br><br>Ive implemented the classical MultiDimensional Scaling for the scikit learn using both functions. Their behavior surprised me for &quot;big&quot; arrays (10000 by 10000, symmetric as it is a similarity matrix).<br>
linalg.svd() raises a memory error because it tries to allocate a (7000000,) array (in fact bigger than that !). This is strange because the test was made on a 64bits Linux, so memory should not have been a problem.<br>linalg.eigh() fails to diagonalize the matrix, it gives me NaN as a result, and this is not very useful.<br>
A direct optimization of the underlying cost function can give me an adequate solution.<br><br>I cannot attach the matrix file (more than 700MB when pickled), but if anyone has a clue, I&#39;ll be glad.<br><br clear="all">
Matthieu<br>-- <br>French PhD student<br>Website : <a href="http://matthieu-brucher.developpez.com/">http://matthieu-brucher.developpez.com/</a><br>Blogs : <a href="http://matt.eifelle.com">http://matt.eifelle.com</a> and <a href="http://blog.developpez.com/?blog=92">http://blog.developpez.com/?blog=92</a><br>
LinkedIn : <a href="http://www.linkedin.com/in/matthieubrucher">http://www.linkedin.com/in/matthieubrucher</a>