Irrelevant Environment Variables causing hash problems #368
Labels
Comments
bpinsard
pushed a commit
to bpinsard/nipype
that referenced
this issue
May 2, 2012
* master: (21 commits) doc: updated debug config message. ants code cleanup (PEP8, unused variables) removed in_file from DTLUTGen docstring added changelog message removed unused variables removed unused imports pep8 and reindent removed an evil trailing comma from WIMT fix: added test to check that inputs are not modified when setting display variable via config fix: added tests, actually added display variable to config fix: interface environment always overwrites config environ DISPLAY doc: change closes nipy#368 fix: environment inputs not changed by config display variable AFNITraitedSpec for Calc Renamed NBS interface Threshold to a float Typo fix Replaced CFF option with multiple network output Added optional node position network Initial NBS commit ...
satra
added a commit
to satra/nipype
that referenced
this issue
May 4, 2012
* upstream/master: fixed cleanup bug more tests: subdirs, files not listed in outputs, removed_unnecessary_outputs/keep_inputs interaction grammar Explained the files removal feature better. A test case for output removal doc: updated debug config message. ants code cleanup (PEP8, unused variables) removed in_file from DTLUTGen docstring fix: added test to check that inputs are not modified when setting display variable via config fix: added tests, actually added display variable to config fix: interface environment always overwrites config environ DISPLAY doc: change closes nipy#368 fix: environment inputs not changed by config display variable Fixes to fsl interface out_file fields
satra
added a commit
to satra/nipype
that referenced
this issue
May 4, 2012
* master: doc: updated debug config message. ants code cleanup (PEP8, unused variables) removed in_file from DTLUTGen docstring fix: added test to check that inputs are not modified when setting display variable via config fix: added tests, actually added display variable to config fix: interface environment always overwrites config environ DISPLAY doc: change closes nipy#368 fix: environment inputs not changed by config display variable Fixes to fsl interface out_file fields
bpinsard
pushed a commit
to bpinsard/nipype
that referenced
this issue
Jul 11, 2012
* master: (31 commits) doc: updated debug config message. ants code cleanup (PEP8, unused variables) removed in_file from DTLUTGen docstring added changelog message removed unused variables removed unused imports pep8 and reindent removed an evil trailing comma from WIMT fix: added test to check that inputs are not modified when setting display variable via config fix: added tests, actually added display variable to config fix: interface environment always overwrites config environ DISPLAY doc: change closes nipy#368 fix: environment inputs not changed by config display variable AFNITraitedSpec for Calc this is the nth time that I change this... test/fixs fixing what I messed up fixing abspath multiple files for tcat and merge default suffix more appropriated ...
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Changing the environment variable DISPLAY will cause nodes to begin to rerun unnecessarily.
The 'environ' input claims to have differing values:
Error from a multisubject connectivity pipeline study (for a cmtk.Parcellate node):
" values differ in fields: environ: {'DISPLAY': ':1', 'SUBJECTS_DIR': '/media/Mobsol/subjects'} != {u'SUBJECTS_DIR': u'/media/Mobsol/subjects'}"
Removing:
display_variable = :1
from the nipype.cfg file fixed it.
Maybe we can work around this problem with a list of relevant environment variables (e.g. FSLOUTPUTTYPE, FSL_DIR, etc..)?
The text was updated successfully, but these errors were encountered: