def parse_infile_names(self):
"""try to get subject and possibly group names from infiles
fill self.snames and self.gnames, if possible
1. get SID
- if files look like out.ss_review.SID.txt, that is a good start
- else, look for varying part of filename
2. get GID
- replace SID in infile names and for varying group name
"""
rv, slist = UTIL.list_minus_pref_suf(self.infiles,'out.ss_review.','.txt')
if rv < 0: return
if rv > 0:
if self.verb > 1: print '++ trying to get SID from glob form'
slist = UTIL.list_minus_glob_form(self.infiles, strip='dir')
else:
if self.verb > 1: print "++ have SIDs from 'out.ss_reiview' form"
if len(slist) == 0:
if self.verb > 1: print "-- empty SID list"
return
# make sure names are unique and not empty
if not UTIL.vals_are_unique(slist):
if self.verb > 1: print '-- SIDs not detected: not unique'
return
minlen = min([len(ss) for ss in slist])
if minlen < 1:
if self.verb > 1: print '-- SIDs not detected: some would be empty'
return
# we have a subject list
self.snames = slist
# now go for GID, start by replacing SIDs in infiles
newfiles = [fname.replace(slist[ind], 'SUBJ') for ind, fname in
enumerate(self.infiles)]
if UTIL.vals_are_constant(newfiles):
print '-- no groups detected from filenames'
return
# okay, try to make a group list
glist = UTIL.list_minus_glob_form(newfiles)
# cannot have dirs in result
for gid in glist:
if gid.find('/') >= 0:
if self.verb>1: print '-- no GIDs, dirs vary in multiple places'
return
minlen = min([len(ss) for ss in glist])
if minlen < 1:
if self.verb > 1: print '-- GIDs not detected: some would be empty'
return
if self.verb > 1: print "++ have GIDs from infiles"
self.gnames = glist
def parse_infile_names(self):
"""try to get subject and possibly group names from infiles
fill self.snames and self.gnames, if possible
1. get SID
- if files look like out.ss_review.SID.txt, that is a good start
- else, look for varying part of filename
2. get GID
- replace SID in infile names and for varying group name
"""
rv, slist = UTIL.list_minus_pref_suf(self.infiles,'out.ss_review.','.txt')
if rv < 0: return
if rv > 0:
if self.verb > 1: print '++ trying to get SID from glob form'
slist = UTIL.list_minus_glob_form(self.infiles, strip='dir')
else:
if self.verb > 1: print "++ have SIDs from 'out.ss_reiview' form"
if len(slist) == 0:
if self.verb > 1: print "-- empty SID list"
return
# make sure names are unique and not empty
if not UTIL.vals_are_unique(slist):
if self.verb > 1: print '-- SIDs not detected: not unique'
return
minlen = min([len(ss) for ss in slist])
if minlen < 1:
if self.verb > 1: print '-- SIDs not detected: some would be empty'
return
# we have a subject list
self.snames = slist
# now go for GID, start by replacing SIDs in infiles
newfiles = [fname.replace(slist[ind], 'SUBJ') for ind, fname in
enumerate(self.infiles)]
if UTIL.vals_are_constant(newfiles):
print '-- no groups detected from filenames'
return
# okay, try to make a group list
glist = UTIL.list_minus_glob_form(newfiles)
# cannot have dirs in result
for gid in glist:
if gid.find('/') >= 0:
if self.verb>1: print '-- no GIDs, dirs vary in multiple places'
return
minlen = min([len(ss) for ss in glist])
if minlen < 1:
if self.verb > 1: print '-- GIDs not detected: some would be empty'
return
if self.verb > 1: print "++ have GIDs from infiles"
self.gnames = glist
def set_ids_from_dsets(self, prefix='', suffix='', hpad=0, tpad=0, dpre=0):
"""use the varying part of the dataset names for subject IDs
If hpad > 0 or tpad > 0, expand into the head or tail of the dsets.
If prefix or suffix is passed, apply them.
return 0 on success, 1 on error
"""
if hpad < 0 or tpad < 0:
print('** set_ids_from_dsets: will not apply negative padding')
return 1
# try filenames without paths, first
dlist = [s.dset.split('/')[-1] for s in self.subjects]
if UTIL.vals_are_constant(dlist):
print('** constant dataset names (%s)' % dlist[0])
print(' trying directories...')
dlist = [s.dset for s in self.subjects]
slist = UTIL.list_minus_glob_form(dlist,
hpad,
tpad,
keep_dent_pre=dpre)
# in the case of diretories, check for success
# (maybe we can try to skip past them, that might be okay)
for index in range(len(slist)):
if '/' in slist[index]:
posn = slist[index].rfind('/')
slist[index] = slist[index][posn + 1:]
if len(slist[index]) < 1:
print(
'** failed to extract subject IDs from directory list')
print(' (directories do not vary at single level)')
return 1
if len(slist) != len(self.subjects):
print('** failed to set SIDs from dset names\n' \
' dsets = %s\n' \
' slist = %s' % (dlist, slist))
return 1
if not UTIL.vals_are_unique(slist):
print('** cannot set IDs from dsets, labels not unique: %s' %
slist)
print('-- labels come from dsets: %s' % dlist)
return 1
for ind, subj in enumerate(self.subjects):
subj.sid = '%s%s%s' % (prefix, slist[ind], suffix)
return 0
def set_ids_from_dsets(self, prefix='', suffix='', hpad=0, tpad=0, dpre=0):
"""use the varying part of the dataset names for subject IDs
If hpad > 0 or tpad > 0, expand into the head or tail of the dsets.
If prefix or suffix is passed, apply them.
return 0 on success, 1 on error
"""
if hpad < 0 or tpad < 0:
print '** set_ids_from_dsets: will not apply negative padding'
return 1
# try filenames without paths, first
dlist = [s.dset.split('/')[-1] for s in self.subjects]
if UTIL.vals_are_constant(dlist):
print '** constant dataset names (%s)' % dlist[0]
print ' trying directories...'
dlist = [s.dset for s in self.subjects]
slist = UTIL.list_minus_glob_form(dlist, hpad, tpad, keep_dent_pre=dpre)
# in the case of diretories, check for success
# (maybe we can try to skip past them, that might be okay)
for index in range(len(slist)):
if '/' in slist[index]:
posn = slist[index].rfind('/')
slist[index] = slist[index][posn+1:]
if len(slist[index]) < 1:
print '** failed to extract subject IDs from directory list'
print ' (directories do not vary at single level)'
return 1
if len(slist) != len(self.subjects):
print '** failed to set SIDs from dset names\n' \
' dsets = %s\n' \
' slist = %s' % (dlist, slist)
return 1
if not UTIL.vals_are_unique(slist):
print '** cannot set IDs from dsets, labels not unique: %s' % slist
print '-- labels come from dsets: %s' % dlist
return 1
for ind, subj in enumerate(self.subjects):
subj.sid = '%s%s%s' % (prefix, slist[ind], suffix)
return 0