def _data(self):
#notunique = (len([r for r in self.bc.sample_id if r]) !=
#len([k for k in self.bc._byCol__indexes['sample_id']
#if k != 'sample_id' and k]))
sids = Counter(r for r in self.bc.sample_id if r)
notunique = {s:c for s, c in sids.items() if c > 1}
values = list(self)
if notunique:
if not values or 'sample_id' not in values[0]:
# [] [{}], caused by a bunch of N/A or missing rows
breakpoint()
msg = f'sample_ids are not unique for {self.path}\n{notunique}'
logd.critical(msg)
self.addError(msg)
# FIXME this needs to be pipelined so we can do the diff ??
for v in values:
v['local_sample_id'] = v['sample_id']
v['sample_id'] = v['subject_id'] + '-' + v['sample_id'] # FIXME '/' gets quoted ...
out = {self.dict_key: values}
for k, heads in self.horizontals.items():
# TODO make sure we actually check that the horizontal
# isn't used by someone else already ... shouldn't be
# it should be skipped but maybe not?
tups = sorted(set(_ for _ in zip(*(getattr(self.bc, head, [])
# FIXME one [] drops whole horiz group ...
for head in heads))
if any(_)))
if tups:
out[k] = [{k:v for k, v in zip(heads, t) if v} for t in tups]
self.embedErrors(out)
return out
def _data(self):
# sigh
#notunique = (len([r for r in self.bc.subject_id if r]) !=
#len([k for k in self.bc._byCol__indexes['subject_id']
#if k != 'subject_id' and k]))
sids = Counter(r for r in self.bc.subject_id if r)
notunique = {s:c for s, c in sids.items() if c > 1}
values = list(self)
if notunique:
msg = f'subject_ids are not unique for {self.path}\n{notunique}'
logd.critical(msg)
self.addError(msg)
out = {self.dict_key: values}
for k, heads in self.horizontals.items():
# TODO make sure we actually check that the horizontal
# isn't used by someone else already ... shouldn't be
# it should be skipped but maybe not?
tups = sorted(set(_ for _ in zip(*(getattr(self.bc, head, [])
# FIXME one [] drops whole horiz group ...
for head in heads))
if any(_)))
if tups:
out[k] = [{k:v for k, v in zip(heads, t) if v} for t in tups]
self.embedErrors(out)
return out