Hi Yarik,<br><br>For both systems:<br><br>issubclass(LinearCSVMC,libsvm.SVM) <br>is True,<br>and issubclass(LinearCSVMC,sg.SVM)<br>is False.<br><br><br><blockquote style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex" class="gmail_quote">
another thing you could if that was under ipython,<br>
enable debug mode by entering<br><br>
pdb<br><br>
so it would jump into pdb whenever failure occurs... and then share output of<br><br>
print dist_samples.shape<br>
print shape<br></blockquote>
<br>ipdb> print dist_samples.shape<br>(0, 1)<br>ipdb> print shape<br>(0,)<br><br><br><blockquote style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex" class="gmail_quote">could you share complete snippet of your code and the dataset (check --<br>
may be it fails also with a small portion of it, e.g. fds1[:, :10]).<br>
you can easily share it by saving via h5save<br></blockquote><br><br>This is where it gets a little strange: so I saved my fds1 to hdf5 format. When I loaded it back, the classifier actually works (on the non-working system)! So in my script I added a few lines that saved and then loaded the dataset just before applying the classifier -- I obtain exactly the same results as on the working system. So I guess it would not make sense to share the hdf5 file with you -- would you like the .nii instead? And here's the code i used for the classifier:<br>
<br> repeater = Repeater(count=200)<br><br> partitioner=ChainNode([NFoldPartitioner(cvtype=1),<br> Balancer(attr='targets',<br> count=1, # for real data > 1<br>
limit='partitions',<br> apply_selection=True<br> )],<br> space='partitions') <br><br>
clf = LinearCSVMC() #choice of classifier <br> permutator = AttributePermutator('targets', limit={'partitions': 1}, count=1)<br> null_cv = CrossValidation(<br> clf,<br> ChainNode([partitioner, permutator], space=partitioner.get_space()),<br>
errorfx=mean_mismatch_error)<br> distr_est = MCNullDist(repeater, tail='left', measure=null_cv,<br> enable_ca=['dist_samples'])<br><br> cvte = CrossValidation(clf, partitioner,<br> errorfx=mean_mismatch_error,<br>
null_dist=distr_est,<br> enable_ca=['stats'])<br><br><br><br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
also what is the output of<br>
<br>
mvpa2.wtf() # where fails<br><br></blockquote><div><br><br>for the working system:<br><br><br>Current date: 2012-09-07 12:59<br>PyMVPA:<br> Version: 2.1.0<br> Hash: 828f4b4bea2488f8c7e3f6c2446a445d77325338<br>
Path: /usr/lib/pymodules/python2.6/mvpa2/__init__.pyc<br> Version control (GIT):<br> GIT information could not be obtained due "/usr/lib/pymodules/python2.6/mvpa2/.. is not under GIT"<br>SYSTEM:<br> OS: posix Linux 2.6.32-25-generic #45-Ubuntu SMP Sat Oct 16 19:52:42 UTC 2010<br>
Distribution: Ubuntu/10.04/lucid<br>EXTERNALS:<br> Present: atlas_fsl, cPickle, ctypes, good scipy.stats.rv_discrete.ppf, griddata, gzip, h5py, ipython, liblapack.so, libsvm, libsvm verbosity control, lxml, matplotlib, nibabel, nose, numpy, pylab, pylab plottable, pywt, pywt wp reconstruct, reportlab, running ipython env, scipy, sg ge 0.6.4, sg ge 0.6.5, sg_fixedcachesize, shogun, shogun.krr, shogun.mpd, shogun.svmocas, skl, weave<br>
Absent: atlas_pymvpa, cran-energy, elasticnet, glmnet, good scipy.stats.rdist, hcluster, lars, mass, mdp, mdp ge 2.4, nipy, nipy.neurospin, openopt, pprocess, pywt wp reconstruct fixed, rpy2, shogun.lightsvm, shogun.svrlight, statsmodels<br>
Versions of critical externals:<br> shogun:full : 0.9.1_r4481_2009-11-16_10:14_<br> shogun:rev : 4481<br> reportlab : 2.4<br> shogun : 0.9.1<br> nibabel : 1.2.2<br> matplotlib : 0.99.1.1<br> scipy : 0.7.0<br>
ipython : 0.10<br> skl : 0.11<br> numpy : 1.3.0<br> ctypes : 1.1.0<br> matplotlib : 0.99.1.1<br> lxml : 2.2.4<br> nifti : failed to query due to "nifti is not a known dependency key."<br>
numpy : 1.3.0<br> pywt : 0.1.7<br> shogun : v0.9.1_r4481_2009-11-16_10:14_<br> Matplotlib backend: TkAgg<br>RUNTIME:<br> PyMVPA Environment Variables:<br> PyMVPA Runtime Configuration:<br> [externals]<br>
have griddata = yes<br> have pprocess = no<br> have ipython = yes<br> have cran-energy = no<br> have good scipy.stats.rdist = no<br> have pylab plottable = yes<br> have pywt wp reconstruct = yes<br> have ctypes = yes<br>
have lxml = yes<br> have running ipython env = yes<br> have nibabel = yes<br> have sg_fixedcachesize = yes<br> have elasticnet = no<br> have shogun.mpd = yes<br> have matplotlib = yes<br> have pywt wp reconstruct fixed = no<br>
have scipy = yes<br> have reportlab = yes<br> have openopt = no<br> have libsvm = yes<br> have h5py = yes<br> have liblapack.so = yes<br> have weave = yes<br> have atlas_fsl = yes<br> have mdp = no<br> have hcluster = no<br>
have sg ge 0.6.4 = yes<br> have sg ge 0.6.5 = yes<br> have good scipy.stats.rv_discrete.ppf = yes<br> have skl = yes<br> have mdp ge 2.4 = no<br> have shogun.svrlight = no<br> have shogun = yes<br> have shogun.svmocas = yes<br>
have nipy = no<br> have glmnet = no<br> have nose = yes<br> have mass = no<br> have lars = no<br> have libsvm verbosity control = yes<br> have atlas_pymvpa = no<br> have shogun.krr = yes<br> have cpickle = yes<br>
have numpy = yes<br> have pylab = yes<br> have rpy2 = no<br> have shogun.lightsvm = no<br> have pywt = yes<br> have statsmodels = no<br> have gzip = yes<br> have nipy.neurospin = no<br> <br> [general]<br> verbose = 1<br>
Process Information:<br> Name: ipython<br> State: R (running)<br> Tgid: 3040<br> Pid: 3040<br> PPid: 1691<br> TracerPid: 0<br> Uid: 1001 1001 1001 1001<br> Gid: 1001 1001 1001 1001<br>
FDSize: 256<br> Groups: 4 20 21 24 26 29 30 44 46 103 104 112 115 120 1001 <br> VmPeak: 697704 kB<br> VmSize: 614156 kB<br> VmLck: 0 kB<br> VmHWM: 235484 kB<br> VmRSS: 131248 kB<br>
VmData: 118084 kB<br> VmStk: 244 kB<br> VmExe: 2160 kB<br> VmLib: 84924 kB<br> VmPTE: 1156 kB<br> Threads: 2<br> SigQ: 0/16382<br> SigPnd: 0000000000000000<br> ShdPnd: 0000000000000000<br>
SigBlk: 0000000000000000<br> SigIgn: 0000000001001000<br> SigCgt: 0000000180000002<br> CapInh: 0000000000000000<br> CapPrm: 0000000000000000<br> CapEff: 0000000000000000<br> CapBnd: ffffffffffffffff<br>
Cpus_allowed: fff<br> Cpus_allowed_list: 0-11<br> Mems_allowed: 00000000,00000001<br> Mems_allowed_list: 0<br> voluntary_ctxt_switches: 801<br> nonvoluntary_ctxt_switches: 2002<br><br>******************************************<br>
for the non-working system:<br><br>Current date: 2012-09-07 13:00<br>PyMVPA:<br> Version: 2.1.0<br> Hash: 828f4b4bea2488f8c7e3f6c2446a445d77325338<br> Path: /usr/lib/pymodules/python2.7/mvpa2/__init__.py<br>
Version control (GIT):<br> GIT information could not be obtained due "/usr/lib/pymodules/python2.7/mvpa2/.. is not under GIT"<br>SYSTEM:<br> OS: posix Linux 3.2.0-29-generic #46-Ubuntu SMP Fri Jul 27 17:03:23 UTC 2012<br>
Distribution: Ubuntu/12.04/precise<br>EXTERNALS:<br> Present: cPickle, ctypes, good scipy.stats.rdist, good scipy.stats.rv_discrete.ppf, griddata, gzip, h5py, ipython, liblapack.so, libsvm, libsvm verbosity control, lxml, matplotlib, mdp, mdp ge 2.4, nibabel, nose, numpy, pylab, pylab plottable, pywt, pywt wp reconstruct, reportlab, running ipython env, scipy, sg ge 0.6.4, sg ge 0.6.5, sg_fixedcachesize, shogun, shogun.krr, shogun.mpd, shogun.svmocas, skl, weave<br>
Absent: atlas_fsl, atlas_pymvpa, cran-energy, elasticnet, glmnet, hcluster, lars, mass, nipy, nipy.neurospin, openopt, pprocess, pywt wp reconstruct fixed, rpy2, shogun.lightsvm, shogun.svrlight, statsmodels<br> Versions of critical externals:<br>
shogun:full : 1.1.0_02ce3cd_2011-12-12_08:17_<br> shogun:rev : 2941901<br> reportlab : 2.5<br> shogun : 1.1.0<br> nibabel : 1.2.2<br> matplotlib : 1.1.1rc<br> scipy : 0.9.0<br> ipython : 0.12.1<br>
skl : 0.11<br> mdp : 3.3<br> numpy : 1.6.1<br> ctypes : 1.1.0<br> matplotlib : 1.1.1rc<br> lxml : 2.3.2<br> nifti : failed to query due to "nifti is not a known dependency key."<br>
numpy : 1.6.1<br> pywt : 0.2.0<br> shogun : v1.1.0_02ce3cd_2011-12-12_08:17_<br> Matplotlib backend: TkAgg<br>RUNTIME:<br> PyMVPA Environment Variables:<br> PYTHONPATH : ":/home/shim/.python27_compiled:/usr/lib/python2.7/plat-linux2:/usr/bin:/usr/lib/python2.7/dist-packages/ubuntuone-couch:/data/projects/09dec11er:/usr/lib/python2.7/dist-packages:/usr/lib/python2.7/dist-packages/IPython/extensions:/usr/lib/python2.7:/usr/lib/python2.7/dist-packages/PIL:/usr/lib/python2.7/dist-packages/ubuntuone-storage-protocol:/usr/lib/python2.7/lib-tk:/usr/lib/python2.7/lib-dynload:/usr/lib/python2.7/dist-packages/ubuntu-sso-client:/usr/lib/python2.7/dist-packages/ubuntuone-client:/usr/lib/python2.7/dist-packages/ubuntuone-installer:/usr/lib/python2.7/lib-old:/usr/local/lib/python2.7/dist-packages:/usr/lib/python2.7/dist-packages/gtk-2.0:/usr/lib/python2.7/dist-packages/ubuntuone-control-panel:/usr/lib/pymodules/python2.7:/usr/lib/python2.7/dist-packages/gst-0.10"<br>
PyMVPA Runtime Configuration:<br> [general]<br> verbose = 1<br> <br> [externals]<br> have running ipython env = yes<br> have numpy = yes<br> have scipy = yes<br> have matplotlib = yes<br> have h5py = yes<br> have reportlab = yes<br>
have weave = yes<br> have good scipy.stats.rdist = yes<br> have good scipy.stats.rv_discrete.ppf = yes<br> have pylab = yes<br> have lars = no<br> have elasticnet = no<br> have glmnet = no<br> have skl = yes<br> have ctypes = yes<br>
have libsvm = yes<br> have shogun = yes<br> have sg ge 0.6.5 = yes<br> have shogun.mpd = yes<br> have shogun.lightsvm = no<br> have shogun.svrlight = no<br> have shogun.krr = yes<br> have shogun.svmocas = yes<br>
have sg_fixedcachesize = yes<br> have openopt = no<br> have nibabel = yes<br> have mdp = yes<br> have mdp ge 2.4 = yes<br> have statsmodels = no<br> have pywt = yes<br> have cpickle = yes<br> have gzip = yes<br>
have cran-energy = no<br> have griddata = yes<br> have nipy.neurospin = no<br> have lxml = yes<br> have atlas_fsl = no<br> have atlas_pymvpa = no<br> have hcluster = no<br> have ipython = yes<br> have liblapack.so = yes<br>
have libsvm verbosity control = yes<br> have mass = no<br> have nipy = no<br> have nose = yes<br> have pprocess = no<br> have pylab plottable = yes<br> have pywt wp reconstruct = yes<br> have pywt wp reconstruct fixed = no<br>
have rpy2 = no<br> have sg ge 0.6.4 = yes<br> Process Information:<br> Name: python<br> State: R (running)<br> Tgid: 3346<br> Pid: 3346<br> PPid: 2453<br> TracerPid: 0<br> Uid: 1000 1000 1000 1000<br>
Gid: 1000 1000 1000 1000<br> FDSize: 256<br> Groups: 4 24 27 30 46 109 124 1000 <br> VmPeak: 875080 kB<br> VmSize: 793920 kB<br> VmLck: 0 kB<br> VmPin: 0 kB<br> VmHWM: 249916 kB<br>
VmRSS: 145868 kB<br> VmData: 274428 kB<br> VmStk: 264 kB<br> VmExe: 2500 kB<br> VmLib: 85992 kB<br> VmPTE: 1240 kB<br> VmSwap: 0 kB<br> Threads: 3<br> SigQ: 0/127910<br>
SigPnd: 0000000000000000<br> ShdPnd: 0000000000000000<br> SigBlk: 0000000000000000<br> SigIgn: 0000000001001000<br> SigCgt: 0000000180000002<br> CapInh: 0000000000000000<br> CapPrm: 0000000000000000<br>
CapEff: 0000000000000000<br> CapBnd: ffffffffffffffff<br> Cpus_allowed: ff<br> Cpus_allowed_list: 0-7<br> Mems_allowed: 00000000,00000001<br> Mems_allowed_list: 0<br> voluntary_ctxt_switches: 627<br>
nonvoluntary_ctxt_switches: 1414<br> <br></div></div>