pos, read_details = ht.pysam_to_hdf(bam_paths[0])
if not bam_input and config.getboolean('behavior', 'deletebam'):
os.remove(bam_paths[0])
binary = np.sign(pos) # dtype=np.uint16
# dimensionality reduction and typing
alleles_to_keep = list(filter(is_frequent, binary.columns))
binary = binary[alleles_to_keep]
if VERBOSE:
print("\n", ht.now(),
'temporary pruning of identical rows and columns')
unique_col, representing = ht.prune_identical_alleles(binary,
report_groups=True)
representing_df = pd.DataFrame([[a1, a2]
for a1, a_l in representing.items()
for a2 in a_l],
columns=['representative', 'represented'])
temp_pruned = ht.prune_identical_reads(unique_col)
if VERBOSE:
print("\n", ht.now(), 'Size of mtx with unique rows and columns:',
temp_pruned.shape)
print(ht.now(), 'determining minimal set of non-overshadowed alleles')
minimal_alleles = ht.prune_overshadowed_alleles(temp_pruned)
if VERBOSE:
else:
print "\nCould not match paired-end pairs. Switching to single-end pipeline."
binary = binary1
is_paired = False
else:
pos, etc, desc = ht.sam_to_hdf(out_dir+"/"+date+"_0.sam", verbosity=args.verbose)
binary = pos.applymap(bool).applymap(int)
#dimensionality reduction and typing
alleles_to_keep = filter(is_frequent, binary.columns)
binary = binary[alleles_to_keep]
if args.verbose:
print "\n", ht.now(), 'temporary pruning of identical rows and columns'
unique_col, representing = ht.prune_identical_alleles(binary, report_groups=True)
representing_df = pd.DataFrame([[a1, a2] for a1, a_l in representing.iteritems() for a2 in a_l],
columns=['representative', 'represented'])
temp_pruned = ht.prune_identical_reads(unique_col)
if args.verbose:
print "\n", ht.now(), 'Size of mtx with unique rows and columns:', temp_pruned.shape
print ht.now(), 'determining minimal set of non-overshadowed alleles'
minimal_alleles = ht.prune_overshadowed_alleles(temp_pruned)
if args.verbose:
print "\n", ht.now(), 'Keeping only the minimal number of required alleles', minimal_alleles.shape
binary = binary[minimal_alleles]
