def setUpClass(cls):
log.info("Loading alignments from {c}".format(c=cls.BAM_PATH))
movie_names, unrolled, datum, columns = from_alignment_file(
alignment_info_from_bam(cls.BAM_PATH)[cls.MOVIE])
cls.datum = datum
cls.unrolled = unrolled
cls.movie_names = movie_names
cls.columns = columns
def setUpClass(cls):
log.info("Loading alignments from {c}".format(c=cls.BAM_PATH))
movie_names, unrolled, datum, columns = from_alignment_file(
alignment_info_from_bam(cls.BAM_PATH)[cls.MOVIE])
cls.datum = datum
cls.unrolled = unrolled
cls.movie_names = movie_names
cls.columns = columns
def analyze_movies(movies, alignment_file_names, stats_models):
all_results = []
log.info("collecting data from {n} BAM files...".format(
n=len(alignment_file_names)))
for file_name in alignment_file_names:
log.info("reading {f}.pbi".format(f=file_name))
results = alignment_info_from_bam(file_name)
for movie, aln_info in results.iteritems():
log.info("Analyzing Movie {n} in {f}".format(n=movie, f=file_name))
args = from_alignment_file(aln_info)
_process_movie_data(movie, file_name, stats_models, *args)
log.info("Completed analyzing {n} movies.".format(n=len(movies)))
def analyze_movies(movies, alignment_file_names, stats_models):
all_results = []
log.info("collecting data from {n} BAM files...".format(
n=len(alignment_file_names)))
for file_name in alignment_file_names:
log.info("reading {f}.pbi".format(f=file_name))
results = alignment_info_from_bam(file_name)
for movie, aln_info in results.iteritems():
log.info("Analyzing Movie {n} in {f}".format(n=movie, f=file_name))
args = from_alignment_file(aln_info)
_process_movie_data(movie, file_name, stats_models, *args)
log.info("Completed analyzing {n} movies.".format(n=len(movies)))
def analyze_movie(movie, alignment_file, stats_models):
"""
The regions should only correspond to a single Movie
:type movie: Movie
:type stats_models: list
"""
started_at = time.time()
log.info("Analyzing Movie {n}".format(n=movie))
movie_names, unrolled, data_, columns = from_alignment_file(
movie, alignment_file)
if len(data_) == 0:
msg = "Movie '{n}' produced no alignments.".format(n=movie)
log.warn(msg)
return
crunched = CrunchedAlignments(movie_names, unrolled, data_, columns)
log.debug("Movie names from crunched {m}.".format(m=movie_names))
reads = crunched.reads()
# subreads recarray
# ["Length", "Accuracy", "isFirst", "modStart", "isFullSubread", "isMaxSubread"]
subreads = crunched.subreads()
log.info("Movie")
log.info(movie)
log.info(('Number of reads', len(reads)))
log.info(('Number of subreads', len(subreads)))
for model in stats_models:
if model.filter_func(movie):
for aggregator in model.aggregators:
if aggregator.DATA_TYPE == READ_TYPE:
aggregator.apply(reads)
if aggregator.DATA_TYPE == SUBREAD_TYPE:
aggregator.apply(subreads)
else:
log.warn(
"model {m}. Skipping movie {r}".format(m=repr(model), r=movie))
pass
run_time = time.time() - started_at
_d = dict(n=movie, s=run_time)
log.info("Completed analyzing Movie {n} with in {s:.2f} sec.".format(**_d))
def analyze_movie(movie, alignment_file, stats_models):
"""
The regions should only correspond to a single Movie
:type movie: Movie
:type stats_models: list
"""
started_at = time.time()
movie_names, unrolled, data_, columns = from_alignment_file(
movie, alignment_file)
_process_movie_data(movie, alignment_file, stats_models,
movie_names, unrolled, data_, columns)
run_time = time.time() - started_at
_d = dict(n=movie, s=run_time)
log.info("Completed analyzing Movie {n} with in {s:.2f} sec.".format(**_d))
def analyze_movies(movies, alignment_file_names, stats_models, nproc=1):
#pool = None
#if nproc >= 1:
# # XXX I use nproc-1 here because the callback in the main process
# # actually takes up a lot of time
# log.info("Starting pool of {n} processes".format(n=max(1, nproc-1)))
# pool = multiprocessing.Pool(processes=nproc)
for movie in movies:
for file_name in alignment_file_names:
log.info("Analyzing Movie {n}".format(n=movie))
results = from_alignment_file(movie, file_name)
_process_movie_data(movie, file_name, stats_models, *results)
# FIXME need to re-think this
#def __analyze_movie(args):
# return from_alignment_file(*args)
#__callback = functools.partial(_process_movie_data, movie,
# file_name, stats_models)
#pool.apply_async(from_alignment_file, (movie, file_name),
# callback=__callback)
#pool.close()
#pool.join()
log.info("Completed analyzing {n} movies.".format(n=len(movies)))