[pymvpa] Memory issues with example dataset

[Kemal]

Thanks for helping! It would really be nice to get this to work with a
few optimizations.

> I would recommend to run your example (still not sure which one
> particular) with following environment variables

I'm using the dataset available for download here:
http://www.pymvpa.org/files/pymvpa_exampledata.tar.bz2

>
> MVPA_DEBUG=.* MVPA_DEBUG_METRICS=vmem,tbc
>
> so smth like
>
>  MVPA_DEBUG=.* MVPA_DEBUG_METRICS=vmem,tbc python smlr.py 2>&1 | tee smlr-debug.output
>
> on linux -- that would dump output to screen as well to a file (for more
> convinient inspectioin later on)
>
> what you will see is like
> [DBG             ] DBG{VmSize: 127924 kB}:    Registering metric <mvpa.base.verbosity.TraceBack object at 0x23156d0>
>
> Where VmSize would give you information on memory utilization -- just
> watch then where it grows considerably (there will be also shortned
> version of traceback included)

I've done so, and attached the output. (also uploaded here:
http://www.filedropper.com/script-debug) The VMSize never even goes
above 1GB. Here is the command I am trying to run (after performing
the necessary imports)

attrs = SampleAttributes('data/attributes_literal.txt',literallabels=True)
ds = NiftiDataset(samples='data/bold.nii.gz',labels=attrs.labels,chunks=attrs.chunks,labels_map=True,mask='data/mask.nii.gz')
pcamapper=pca.PCAMapper()
pcamapper.train(ds)

What kind of optimizations would you recommend? Thank you.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: script-debug.output
Type: application/octet-stream
Size: 453274 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/pkg-exppsy-pymvpa/attachments/20090626/498b3c72/attachment-0001.obj>