Databases
StackOverflow2013
Postsdbo
POWeka's PCA is taking too long to run
 

Note that there are some explanatory texts on larger screens.

plurals
  1. POWeka's PCA is taking too long to run
    primarykey
    data
    text
    <p>I am trying to use Weka for feature selection using PCA algorithm.</p> <p>My original feature space contains ~9000 attributes, in 2700 samples. <br>I tried to reduce dimensionality of the data using the following code:</p> <pre><code>AttributeSelection selector = new AttributeSelection(); PrincipalComponents pca = new PrincipalComponents(); Ranker ranker = new Ranker(); selector.setEvaluator(pca); selector.setSearch(ranker); Instances instances = SamplesManager.asWekaInstances(trainSet); try { selector.SelectAttributes(instances); return SamplesManager.asSamplesList(selector.reduceDimensionality(instances)); } catch (Exception e ) { ... } </code></pre> <p>However, It did not finish to run within 12 hours. It is stuck in the method <code>selector.SelectAttributes(instances);</code>.</p> <p><strong>My questions are:</strong> Is so long computation time expected for weka's PCA? Or am I using PCA wrongly?</p> <p><strong>If the long run time is expected:</strong> <br>How can I tune the PCA algorithm to run <em>much</em> faster? Can you suggest an alternative? (+ example code how to use it)?</p> <p><strong>If it is not:</strong> <br>What am I doing wrong? How should I invoke PCA using weka and get my reduced dimensionality?</p> <p><strong>Update:</strong> The comments confirms my suspicion that it is taking much more time than expected. <br><strong>I'd like to know</strong>: How can I get PCA in java - using weka or an alternative library. <br>Added a bounty for this one.</p>
    singulars
    1. PO
      singulars
      1. PTAnswer
    1. This table or related slice is empty.
    1. PTQuestion
    1. USamit
    1. USamit
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. PO
      singulars
      1. PTAnswer
    2. PO
      singulars
      1. PTAnswer
    1. VO
      singulars
      1. POWeka's PCA is taking too long to run
      1. USamit
      1. VTBountyStart
    2. VO
      singulars
      1. POWeka's PCA is taking too long to run
      1. This table or related slice is empty.
      1. VTUpMod
    3. VO
      singulars
      1. POWeka's PCA is taking too long to run
      1. USbouteillebleu
      1. VTFavorite
    1. COI don't know weka, but this isn't a reasonable amount of time. PCA takes as long to run as matrix multiplication (because that's what it is, essentially).
      singulars
      1. POWeka's PCA is taking too long to run
      1. USGuy Adini
    2. CO@GuyAdini: I expected it is not reasonable, and I am misusing it. Do you know of an open source library [preferably in java] that I can use? [I'd also be happy to get a sample code how to use it]
      singulars
      1. POWeka's PCA is taking too long to run
      1. USamit
    3. COSorry - never did that in Java (otherwise I'd post an answer). If you want another language, try Python's sklearn. Anyway, coding PCA yourself is about 3 lines of code, given a matrix manipulation library: set S = X*X^t, then take the first k eigenvectors of S.
      singulars
      1. POWeka's PCA is taking too long to run
      1. USGuy Adini
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload