def main(args=None):
# loading taxa names
taxa = load_taxa(args['<taxa>'])
# loading KDEs
KDEs = Utils.load_kde(args['<kde>'])
# parsing KDEs
kde_type = Utils.KDE_type(KDEs)
# parsing KDE
if kde_type == 1:
KDEs_p = [[t, k] for t, k in KDEs if t in taxa]
elif kde_type == 2:
KDEs_p = {t: k for t, k in KDEs.items() if t in taxa}
elif kde_type == 3:
KDEs_p = {}
for libID, v in KDEs_p.items():
KDEs_pp = {t: k for t, k in v.items() if t in taxa}
KDEs_p[libID] = KDEs_pp
KDEs_pp = None
elif kde_type == 4:
KDEs_p = {}
for libID, filename in KDEs.items():
KDE_bylib = Utils.load_kde(filename)
KDE_bylib = {t: k for t, k in KDE_bylib.items() if t in taxa}
KDEs_p[libID] = KDE_bylib
KDE_bylib = None
else:
raise TypeError, 'KDE object type not recognized'
# writing
dill.dump(KDEs_p, sys.stdout)
def main(args=None):
KDEs = Utils.load_kde(args['<kde>'])
# KDE object type
kde_type = Utils.KDE_type(KDEs)
# info for KDEs
print '\t'.join(['libID', 'taxon', 'bandwidth'])
if kde_type == 1:
for t, k in KDEs:
print '\t'.join(['1', t, kde_factor(k)])
elif kde_type == 2:
for t, k in KDEs.items():
print '\t'.join(['1', t, kde_factor(k)])
elif kde_type == 3:
for libID, x in KDEs.items():
for t, k in x.items():
print '\t'.join([libID, t, kde_factor(k)])
elif kde_type == 4:
for libID, filename in KDEs.items():
KDE_bylib = Utils.load_kde(filename)
for t, k in KDE_bylib.items():
print '\t'.join([libID, t, kde_factor(k)])
else:
raise TypeError, 'KDE object type not recognized'
def main(args=None):
# load KDEs object
sys.stderr.write('Loading KDEs...\n')
KDEs = Utils.load_kde(args['<kde>'])
# header
if args['-s']:
print '\t'.join(['lib_ID', 'taxon_ID', 'KDE_ID', 'min', 'percentile_5',
'percentile_25', 'mean', 'median', 'percentile_75',
'percentile_95', 'max', 'stdev'])
# KDE type
kde_type = Utils.KDE_type(KDEs)
# parsing KDE
if kde_type == 1:
if args['-n']:
print len(KDEs)
sys.exit()
for x in KDEs:
if args['-s']:
KDE_dataset_stats(x[1], x[0])
else:
print x[0]
elif kde_type == 2:
if args['-n']:
print len(KDEs.keys())
sys.exit()
for x,y in KDEs.items():
if args['-s']:
KDE_dataset_stats(y, x)
else:
print x
elif kde_type == 3:
if args['-n']:
print '\t'.join(['library', 'N'])
for libID,v in KDEs.items():
print '\t'.join([str(x) for x in [libID, len(v.keys())]])
sys.exit()
for x,y in KDEs.items():
for xx,yy in y.items():
if args['-s']:
KDE_dataset_stats(yy, xx, libID=x)
else:
print '\t'.join([x,xx])
elif kde_type == 4:
for libID,filename in KDEs.items():
KDE_bylib = Utils.load_kde(filename)
if args['-n']:
print len(KDE_bylib.keys())
sys.exit()
for x,y in KDE_bylib.items():
if args['-s']:
KDE_dataset_stats(y, x)
else:
print x
else:
raise TypeError, 'KDE object type not recognized'
def main(args=None):
# input
KDEs = Utils.load_kde(args['<kde>'])
n = int(args['-n'])
# KDE object type
kde_type = Utils.KDE_type(KDEs)
# sampling from KDEs
vals = {}
if kde_type == 1:
vals['1'] = {taxon:kde.resample(n)[0,] for taxon,kde in KDEs \
if kde is not None}
elif kde_type == 2:
vals['1'] = {taxon:kde.resample(n)[0,] for taxon,kde in KDEs.items() \
if kde is not None}
elif kde_type == 3:
for x, y in KDEs.items():
vals[x] = {taxon:kde.resample(n)[0,] for taxon,kde in y.items() \
if kde is not None}
elif kde_type == 4:
for libID, filename in KDEs.items():
KDE_bylib = Utils.load_kde(filename)
vals[libID] = {
taxon: kde.resample(n)[0, ]
for taxon, kde in KDE_bylib.items() if kde is not None
}
else:
raise TypeError, 'KDE object type not recognized'
# writing out results
tbl = pd.concat([to_df(y, x) for x, y in vals.items()])
tbl.to_csv(sys.stdout, sep='\t', index=False)
def test_kde():
f = os.path.join(data_dir, 'ampFrag_skewN90-25-n5-nS_kde.pkl')
# load kde
kde = Utils.load_kde(f)
assert isinstance(kde, dict)
# check type
kde_type = Utils.KDE_type(kde)
assert kde_type == 2
def test_file_status():
# real file
f = os.path.join(data_dir, 'ampFrag_skewN90-25-n5-nS_kde.pkl')
Utils.is_file(f)
Utils.checkExists(f)
Utils.checkEmpty(f)
# fake file
f = os.path.join(data_dir, 'DOES_NOT_EXIST')
with pytest.raises(IOError):
Utils.is_file(f)
with pytest.raises(IOError):
Utils.checkExists(f)
def main(args=None):
args['--walk'] = int(args['--walk'])
# dist params as dict
args['--dist_params'] = Utils.parseKeyValueString(args['--dist_params'])
otu_tbl = OTU_table.from_csv(args['<OTU_table_file>'], sep='\t')
# if --samp_min, get min comm size, set dist to uniform w/ same low & high
if args['--samp_min']:
min_size = otu_tbl.get_comm_size_stats()[0]
assert min_size > 0, '--samp min is < 1. Nothing to sample!'
args['--dist_params'] = {'low': min_size, 'high': min_size}
# setting subsampling size distribution
otu_tbl.set_samp_dist(samp_dist=args['--dist'],
samp_dist_params=args['--dist_params'])
# subsampling
df = otu_tbl.subsample(no_replace=args['--no-replace'],
walk=args['--walk'],
min_size=args['--min_size'],
max_size=args['--max_size'],
base=args['--base'])
# writing out table
df.to_csv(sys.stdout, sep='\t', index=False)
def main(args=None):
for param in [
'birth_rate', 'death_rate', 'birth_rate_sd', 'death_rate_sd'
]:
param = '--' + param
args[param] = float(args[param])
# loading taxon list
if args['<genome_list>'] is not None:
taxa = Utils.parseGenomeList(args['<genome_list>'], check_exists=False)
taxa = [x[0] for x in taxa]
elif args['<comm_file>'] is not None:
comm = CommTable.from_csv(args['<comm_file>'], sep='\t')
taxa = comm.get_unique_taxon_names()
# init dendropy taxon namespace
taxa = dendropy.TaxonNamespace(taxa, label='taxa')
# simulating tree
if args['--star']:
tree = star_tree(taxon_set=taxa)
else:
tree = birth_death(args['--birth_rate'],
args['--death_rate'],
birth_rate_sd=args['--birth_rate_sd'],
death_rate_sd=args['--death_rate_sd'],
num_extant_tips=len(taxa))
# writing tree
outfmt = args['--outfmt'].lower()
psbl_fmts = ['newick', 'nexus']
assert outfmt in psbl_fmts, 'output file format not recognized.' +\
' Possible formats: {}'.format(', '.join(psbl_fmts))
tree.write_to_stream(sys.stdout, outfmt)
def get_taxa(KDEs, kde_type, all_taxa=None):
"""Getting taxa names from KDE object.
"""
# parsing KDE
if kde_type == 1:
if all_taxa is not None:
taxa = [k[0] for k in KDEs]
else:
taxa = [k[0] for k in KDEs if k[1] is not None]
elif kde_type == 2:
if all_taxa is not None:
taxa = KDEs.keys()
else:
taxa = [k for k,v in KDEs.items() if v is not None]
elif kde_type == 3:
taxa = []
for libID,v in KDEs.items():
if all_taxa is not None:
taxa += v.keys()
else:
taxa += [k for k,vv in v.items() if v is not None]
taxa = list(set(taxa))
elif kde_type == 4:
taxa = []
for libID,filename in KDEs.items():
KDE_bylib = Utils.load_kde(filename)
if all_taxa is not None:
taxa += KDE_bylib.keys()
else:
taxa += [k for k,v in KDE_bylib.items() if v is not None]
else:
raise TypeError, 'KDE object type not recognized'
return taxa
def main(args=None):
# subsample with replacement arg
#if args['-r'] is not None:
# args['-r'] = True
#else:
# args['-r'] = False
# load KDEs object
sys.stderr.write('Loading KDEs...\n')
KDEs = Utils.load_kde(args['<kde>'])
# KDE type
kde_type = Utils.KDE_type(KDEs)
# all taxa names in KDEs
taxa = get_taxa(KDEs, kde_type, args['-a'])
ntaxa = len(taxa)
# subsampling (if needed)
## number to subsample
nsub = None
if args['-s'] is not None:
nsub = int(args['-s'])
elif args['-f'] is not None:
nsub = float(args['-f']) * ntaxa
nsub = int(nsub)
elif args['-p'] is not None:
nsub = float(args['-p']) / 100 * ntaxa
nsub = int(nsub)
## subsampling
if nsub is not None:
if nsub > ntaxa:
args['-r'] = True
msg = 'WARNING: nsub > ntaxa, sub-sampling with replacement!'
sys.stderr.write(msg + '\n')
taxa = np.random.choice(taxa, size=nsub, replace=args['-r'])
msg = 'Subsampled {} taxa'
sys.stderr.write(msg.format(nsub) + '\n')
# writing to STDOUT
for x in taxa:
print x