def task_run_modules(ana_pk, **kwargs):
"""run all enabled modules for an analysis
:type ana_pk: Analysis
:param ana_pk: Analysis entity
:keyword: any, will be passed to modules as parameters
:returns: True on success, False on error
"""
ana = None
success = None
logger = Logger.get_logger(StringIO(""))
try:
# get analysis
ana = Analysis.objects.get(pk=ana_pk.pk)
logger.log_delimiter_line()
logger.log("processing %s" % ana)
ana.status = Analysis.STATUS.running
ana.save()
# get module list
mod_list = ana.datafile.dataset.module_set.all()
assert mod_list, "module list is empty!"
# get file handles
logger.log("reading input files")
rd_file = ana.datafile.rd_file
gt_file = ana.datafile.gt_file
st_file = ana.st_file
rd, sampling_rate = read_hdf5_arc(rd_file.data.path)
if sampling_rate is not None:
kwargs.update(sampling_rate=sampling_rate)
ev_sts = read_gdf_sts(st_file.data.path)
gt_sts = read_gdf_sts(gt_file.data.path)
logger.log("done reading input files")
# apply modules
_tick_all = datetime.now()
for mod in mod_list:
logger.log_delimiter_line()
logger.log("starting {}".format(mod))
module_cls = mod.get_module_cls()
_tick_ = datetime.now()
module = module_cls(rd, gt_sts, ev_sts, logger, **kwargs)
module.apply()
module.save(mod, ana)
_tock_ = datetime.now()
logger.log("finished in {}".format(str(_tock_ - _tick_)))
del module, module_cls, _tick_, _tock_
_tock_all = datetime.now()
logger.log_delimiter_line()
logger.log("finished all module in {}".format(str(_tock_all - _tick_all)))
except Exception, ex:
logger.log_delimiter_line()
logger.log("ERROR ({}) :: {}".format(ex.__class__.__name__, ex))
success = False
def task_run_modules(ev, **kwargs):
"""core function to run all modules for an evaluation
:type ev: Evaluation
:param ev: Evaluation entity
:keyword: any, will be passed to modules as parameters
:returns: True on success, False on error
"""
success = None
try:
ev.status = ev.STATUS.running
ev.save()
mod_list = ev.trial.benchmark.module_set.all()
logger = Logger.get_logger(StringIO())
except:
success = False
else:
try:
logger.log_delimiter_line()
rd_file = ev.trial.rd_file
gt_file = ev.trial.gt_file
ev_file = ev.ev_file
logger.log('processing: %s' % ev)
logger.log('reading input files')
rd, sampling_rate = read_hdf5_arc(rd_file.file.path)
if sampling_rate is not None:
kwargs.update(sampling_rate=sampling_rate)
ev_sts = read_gdf_sts(ev_file.file.path)
gt_sts = read_gdf_sts(gt_file.file.path)
logger.log('done reading input files')
logger.log_delimiter_line()
# modules
assert len(mod_list), 'Module list is empty!'
for mod in mod_list:
logger.log('starting module: %s' % mod)
module_pkg = importlib.import_module('spike.module.%s' % mod.path)
_tick_ = datetime.now()
module = module_pkg.module_cls(rd, gt_sts, ev_sts, logger, **kwargs)
module.apply()
module.save(mod, ev)
_tock_ = datetime.now()
logger.log('finished: %s' % str(_tock_ - _tick_))
logger.log_delimiter_line()
del module, module_pkg
except Exception, ex:
logger.log('ERROR: (%s) %s' % (ex.__class__.__name__, str(ex)))
success = False
ev.status = ev.STATUS.failure
else:
def task_validate_spiketrain_file(pk, **kwargs):
"""validate spike train file - that is a text file in gdf format (space separated, 2col, [key,time])
:type pk: int
:param pk: pk of `Datafile` entity
:returns: bool -- True if file validates, False else. Processing
log, including errors, will be written to the `Datafile` entity.
"""
# init and checks
valid = False
logger = Logger.get_logger(StringIO())
try:
obj = Asset.objects.get(id=pk)
assert obj.kind == "st_file"
df = obj.content_object
except:
logger.log("ERROR")
return valid
try:
logger.log("looking at spike train file with pk: %s" % obj.id)
sts = read_gdf_sts(obj.data.path)
logger.log("found st_file: %s" % obj.name)
for st in sts:
if not isinstance(sts[st], sp.ndarray):
raise TypeError("spike train %s not ndarray" % st)
if not sts[st].ndim == 1:
raise ValueError("spike trains have to be ndim==1")
# TODO: more checks?
logger.log("st_file passed all checks")
valid = True
except Exception, ex:
logger.log('ERROR: spike train file check: %s' % str(ex))
def task_validate_spiketrain_file(pk, **kwargs):
"""validate spike train file - that is a text file in gdf format (space separated, 2col, [key,time])
:type pk: int
:param pk: pk of `Datafile` entity
:returns: bool -- True if file validates, False else. Processing
log, including errors, will be written to the `Datafile` entity.
"""
# init and checks
valid = False
logger = Logger.get_logger(StringIO())
try:
obj = Asset.objects.get(id=pk)
assert obj.kind == "st_file"
df = obj.content_object
except:
logger.log("ERROR")
return valid
try:
logger.log("looking at spike train file with pk: %s" % obj.id)
sts = read_gdf_sts(obj.data.path)
logger.log("found st_file: %s" % obj.name)
for st in sts:
if not isinstance(sts[st], sp.ndarray):
raise TypeError("spike train %s not ndarray" % st)
if not sts[st].ndim == 1:
raise ValueError("spike trains have to be ndim==1")
# TODO: more checks?
logger.log("st_file passed all checks")
valid = True
except Exception, ex:
logger.log('ERROR: spike train file check: %s' % str(ex))
def task_validate_spiketrain_file(pk):
"""validate a spiketrain file - that is a text file in gdf format (space separated, 2col, [key,time])
:type pk: int
:param pk: pk for Data entity
:returns: bool -- True if Data validates, False else. Processing
log, including errors, will be written to the Data entity.
"""
# init and checks
valid = False
logger = Logger.get_logger(StringIO())
try:
df = Data.objects.get(id=pk)
assert df.kind == 'st_file'
tr = df.content_object
except:
logger.log('ERROR')
return valid
try:
logger.log('looking at spiketrain file with pk: %s' % df.id)
sts = read_gdf_sts(df.file.path)
logger.log('found st_file: %s' % df.name)
for st in sts:
if not isinstance(sts[st], sp.ndarray):
raise TypeError('spike train %s not ndarray' % st)
if not sts[st].ndim == 1:
raise ValueError('spike trains have to be ndim==1')
# TODO: more checks?
logger.log('st_file passed all checks')
valid = True
except Exception, ex:
logger.log('ERROR: spiketrain file check: %s' % str(ex))
