Commits on Jul 21, 2016
  1. @yarikoptic

    Merge pull request #483 from yarikoptic/enh-part2

    ENH: random and equidistant strategies for FactorialPartitioner
    yarikoptic committed on GitHub Jul 21, 2016
Commits on Jul 20, 2016
  1. @yarikoptic

    ENH+BF: xrandom_iterprod - naive but functional implementation to get…

    … random products among sequences
    
    to make random selection in FactorialPartitioner usable
    yarikoptic committed Jul 20, 2016
  2. @yarikoptic
  3. @yarikoptic
  4. @yarikoptic
Commits on Jul 12, 2016
  1. @yarikoptic

    Merge pull request #482 from nno/_enh/prep_afni_surf_error_if_oblique

    NF: in pymvpa2-prep-afni-afni, throw an error when experimental volume is oblique
    yarikoptic committed on GitHub Jul 12, 2016
  2. @nno

    PEP8: follow formatting guidelines

    nno committed Jul 12, 2016
Commits on Jul 11, 2016
  1. @nno
  2. @nno
  3. @nno
  4. @nno

    SML: remove unnecessary newline

    nno committed Jul 11, 2016
  5. @nno

    BF: Use "=" instead of "is"

    nno committed Jul 11, 2016
  6. @nno

    PEP8: reformatting

    nno committed Jul 11, 2016
  7. @nno

    NF: in pymvpa2-prep-afni-afni, throw an error when the volume to be a…

    …ligned is oblique, because this does not work well together with AFNI's alignment pipeline
    nno committed Jul 11, 2016
Commits on Jun 22, 2016
  1. @yarikoptic
Commits on Jun 15, 2016
  1. @nno

    Merge pull request #477 from feilong/master

    BF: Minor fix and improvement with computing ico linear divisions
    nno committed on GitHub Jun 15, 2016
  2. @feilong

    FIX: minor bug with computing ico linear divisions & minor improvment…

    … with determining multiples
    feilong committed Jun 15, 2016
Commits on Jun 8, 2016
  1. @yarikoptic

    Merge pull request #473 from nno/_doc/surface_qe_args

    DOC: update documentation to properly reflect supported input type
    yarikoptic committed Jun 8, 2016
Commits on May 27, 2016
  1. @nno

    Merge pull request #475 from effigies/mask

    ENH: Permit 4D image mask in fmri_dataset if time dimension is 1
    nno committed May 27, 2016
  2. @effigies

    TEST: Test 4d masks

    effigies committed May 26, 2016
Commits on May 26, 2016
  1. @effigies
Commits on May 21, 2016
  1. @yarikoptic

    Merge pull request #474 from swaroopgj/master

    BF: Relaxed tests a little to account for noisy/correlated features
    Closes #455
    yarikoptic committed May 20, 2016
Commits on May 20, 2016
  1. @swaroopgj

    BF: Relaxed tests a little to account for noisy and correlated featur…

    …es in randomly generated test datasets.
    swaroopgj committed May 20, 2016
  2. @nno
Commits on May 12, 2016
  1. @yarikoptic
Commits on May 11, 2016
  1. @yarikoptic
Commits on May 10, 2016
  1. @yarikoptic

    Merge pull request #468 from yarikoptic/bf-py3

    ENH+BF(PY3): variety of fixes for PY3 (daily tests), and display also externals version
    yarikoptic committed May 10, 2016
  2. @yarikoptic
  3. @yarikoptic
Commits on May 3, 2016
  1. @yarikoptic

    BF: delay import of sklearn until used.

    Otherwise we would need to condition entire module on its presence
    yarikoptic committed May 2, 2016
  2. @yarikoptic

    ENH+BF(PY3): variety of fixes for PY3 (daily tests), and display also…

    … externals version in debug msg if found
    yarikoptic committed May 2, 2016
Commits on Apr 25, 2016
  1. @yarikoptic

    Merge pull request #445 from mvdoc/lassoregr

    ENH,TST: add rsa.Regression measure with tests + small fix on typos [reopens #420]
    yarikoptic committed Apr 25, 2016
  2. @mvdoc

    Merge remote-tracking branch 'origin/master' into lassoregr

    * origin/master: (117 commits)
      ENH: fixed up tests - now adhoc sl should work just fine with CustomPartitioner
      ENH: simplify SplitRFE doc example to nore require errorfx + reorder mappers to have mean(abs(normed)))
      BF: reduce required ACC while testing auc -- ran 100 times without failure
      BF: two tests must be skipped if hdf5 is NA
      DOC: Changed volume geometry warning to print dataset instead of dumping it on console
      minor change as per @nno's suggestion
      OPT: show only summary of representation of internal _src2aux of VolumeMaskDictionary object, as the full output may result in a significant slow-down of the __repr__ method
      DOC: update matrix plot range to clear confusion, and fixed a few typos
      ENH: made tests less stringent and some PEP8 cleanup
      Added a case to test to show it fails for certain cases which will be fixed by this commit
      BF: This handles surface-to-voxel queryengines where setting center seed is not meaningful
      BF: initializing wrong shapes of mappers_full in featselhyper
      BF: making featsel param when given as proportion to behave as intended
      ENH: use "super" when calling __init__ in pdist-related Measures
      BF: changing default nblocks to None
      ENH: Minor change to let the algorithm use nblocks even for nproc=1
      DOC: rudimentary docstring for FeatureSelectionHyperalignment pointing to SearchlightHyperalignment
      Added reference to the paper into ChangeLog
      RF: HyperalignmentMeasure -> FeatureSelectionHyperalignment + minor cleanup
      Fixed typo
      ...
    mvdoc committed Apr 25, 2016
  3. @yarikoptic

    Merge branch 'pr-452'

    * pr-452:
      BF: Fixing hyperalignment example so that it doesn't crash without lapack library
    yarikoptic committed Apr 25, 2016
  4. @yarikoptic

    Merge branch 'master' into pr-272

    * master: (107 commits)
      ENH: simplify SplitRFE doc example to nore require errorfx + reorder mappers to have mean(abs(normed)))
      BF: reduce required ACC while testing auc -- ran 100 times without failure
      BF: two tests must be skipped if hdf5 is NA
      DOC: Changed volume geometry warning to print dataset instead of dumping it on console
      minor change as per @nno's suggestion
      OPT: show only summary of representation of internal _src2aux of VolumeMaskDictionary object, as the full output may result in a significant slow-down of the __repr__ method
      DOC: update matrix plot range to clear confusion, and fixed a few typos
      ENH: made tests less stringent and some PEP8 cleanup
      Added a case to test to show it fails for certain cases which will be fixed by this commit
      BF: This handles surface-to-voxel queryengines where setting center seed is not meaningful
      BF: initializing wrong shapes of mappers_full in featselhyper
      BF: making featsel param when given as proportion to behave as intended
      ENH: use "super" when calling __init__ in pdist-related Measures
      BF: changing default nblocks to None
      ENH: Minor change to let the algorithm use nblocks even for nproc=1
      DOC: rudimentary docstring for FeatureSelectionHyperalignment pointing to SearchlightHyperalignment
      Added reference to the paper into ChangeLog
      RF: HyperalignmentMeasure -> FeatureSelectionHyperalignment + minor cleanup
      Fixed typo
      DOC: added brief changelog entry for slhyper, aiming for 2.5.0 release
      ...
    yarikoptic committed Apr 25, 2016