def calc_stats(file_in,
return_type=None,
txt_width=0,
log=False,
backend=None,
format='fasta',
minLength=0,
**kwargs):
"""
Given contigs in fastsa format:
* calculate length stats (including N50)
* plot histogram (use txt_width and backend to select format)
* return_types:
None: just print text to STDOUT
'report': return text
'data': return dictionary of data
"""
with open(file_in, 'r') as seq:
sizes = [
len(record) for record in SeqIO.parse(seq, format)
if len(record) >= minLength
]
sizes = numpy.array(sizes)
data = get_contig_length_stats(sizes)
if return_type != 'data':
report = get_contig_length_report(data)
if backend is not None:
h = plot_assembly(sizes,
file_in,
data,
log=log,
backend=backend,
**kwargs)
if txt_width > 0:
if backend is None:
h = numpy.histogram(sizes, **kwargs)
histogramText = ascii_histogram(h, log=log, width=txt_width)
if return_type != 'data':
if log:
report += "\n\nContig length histogram (log):\n"
else:
report += "\n\nContig length histogram:\n"
report += histogramText
else:
data['histogram'] = histogramText
if return_type == 'data':
return data
elif return_type is None:
print(report)
else:
return report
def contig_length_stats(contig_stats,
return_type=None,
txt_width=0,
log=False,
min_length=0,
**kwargs):
"""
Given contig stats table
* calculate length stats (including N50)
* optionally plot histogram (use txt_width and backend to select format)
(if txt_width is greater than 0 (should be at least 40 for a good plot))
* return_types:
None: just print text to STDOUT
'report': return text
'data': return dictionary of data
"""
report_data = {"min_length": min_length}
contig_stats = contig_stats.loc[contig_stats.length >= min_length]
if contig_stats.shape[0] == 0:
report_data['Assembly'] = {'count': 0}
else:
report_data['Assembly'] = get_N_stats(contig_stats)
for column, label in {
'length': 'Contig Lengths',
'read count': 'Reads per Contig',
'av cov': 'Mean Mapped Depth',
'mx cov': 'Maximum Mapped Depth',
'mn cov': 'Minimum Mapped Depth',
'GC': 'GC Content'
}.items():
if column not in contig_stats.columns:
continue
report_data[label] = get_column_stats(contig_stats[column])
if txt_width > 0:
report_data[label]['log'] = "(log)" if log else ""
report_data[label]['histogram'] = \
ascii_histogram(numpy.histogram(contig_stats[column],
**kwargs),
log=log,
width=txt_width)
if return_type == 'data':
return report_data
report = get_contig_stats_report(report_data)
if return_type is None:
print(report)
else:
return report
def calc_stats(file_in,
return_type=None,
txt_width=0,
log=False,
backend=None,
format='fasta',
minLength=0,
**kwargs):
"""
Given contigs in fastsa format:
* calculate length stats (including N50)
* plot histogram (use txt_width and backend to select format)
* return_types:
None: just print text to STDOUT
'report': return text
'data': return dictionary of data
"""
with open(file_in, 'r') as seq:
sizes = [len(record) for record in SeqIO.parse(
seq, format) if len(record) >= minLength]
sizes = numpy.array(sizes)
data = get_contig_length_stats(sizes)
if return_type != 'data':
report = get_contig_length_report(data)
if backend is not None:
h = plot_assembly(sizes, file_in, data, log=log,
backend=backend, **kwargs)
if txt_width > 0:
if backend is None:
h = numpy.histogram(sizes, **kwargs)
histogramText = ascii_histogram(h, log=log, width=txt_width)
if return_type != 'data':
if log:
report += "\n\nContig length histogram (log):\n"
else:
report += "\n\nContig length histogram:\n"
report += histogramText
else:
data['histogram'] = histogramText
if return_type == 'data':
return data
elif return_type is None:
print(report)
else:
return report
def contig_length_stats(contig_stats, return_type=None,
txt_width=0,
log=False,
min_length=0,
**kwargs):
"""
Given contig stats table
* calculate length stats (including N50)
* optionally plot histogram (use txt_width and backend to select format)
(if txt_width is greater than 0 (should be at least 40 for a good plot))
* return_types:
None: just print text to STDOUT
'report': return text
'data': return dictionary of data
"""
report_data = {"min_length": min_length}
contig_stats = contig_stats.loc[contig_stats.length >= min_length]
if contig_stats.shape[0] == 0:
report_data['Assembly'] = {'count': 0}
else:
report_data['Assembly'] = get_N_stats(contig_stats)
for column, label in {'length': 'Contig Lengths',
'read count': 'Reads per Contig',
'av cov': 'Mean Mapped Depth',
'mx cov': 'Maximum Mapped Depth',
'mn cov': 'Minimum Mapped Depth',
'GC': 'GC Content'}.items():
if column not in contig_stats.columns:
continue
report_data[label] = get_column_stats(contig_stats[column])
if txt_width > 0:
report_data[label]['log'] = "(log)" if log else ""
report_data[label]['histogram'] = \
ascii_histogram(numpy.histogram(contig_stats[column],
**kwargs),
log=log,
width=txt_width)
if return_type == 'data':
return report_data
report = get_contig_stats_report(report_data)
if return_type is None:
print(report)
else:
return report
def main():
"""
Sets up the command line interface
"""
description = __doc__
parser = argparse.ArgumentParser(description=description)
parser.add_argument('infile',
nargs='?',
type=argparse.FileType('rU'),
default=sys.stdin,
help=("Input file (defaults to STDIN) containing "
"a value on each line"))
parser.add_argument('outfile',
nargs='?',
type=argparse.FileType('w'),
default=sys.stdout,
help="File to which to write histogram")
parser.add_argument('-b', '--bins', type=int, default=50)
parser.add_argument('-l', '--label', default="value")
parser.add_argument('-w', '--width', default=75, type=int)
parser.add_argument('-L', '--log', action='store_true')
parser.add_argument('-W', '--max-label-width', type=int, default=10)
# log level and help
add_universal_arguments(parser)
arguments = parser.parse_args()
setup_logging(arguments)
values = []
for line in arguments.infile:
try:
values.append(float(line.strip()))
except ValueError:
if len(line.strip()) > 0:
logging.warning("Skipping bad line:\n%s", line.strip())
logging.info("Read in %d values", len(values))
arguments.outfile.write(
ascii_histogram(
histogram(values, bins=arguments.bins),
label=arguments.label,
width=arguments.width,
log=arguments.log,
maxLabelWidth=arguments.max_label_width,
))
def mira_stats(contigStatsFile, minLength=0, bins=20, **kwargs):
"""
Get length, coverage, and GC stats from mira info file
Returns text with N50 and histograms
"""
contigStats = pandas.read_csv(contigStatsFile, index_col=0, sep='\t')
if minLength > 0:
contigStats = contigStats[contigStats.length >= minLength]
sizes = contigStats['length']
data = get_contig_length_stats(sizes)
report = get_contig_length_report(data)
# add histograms to report
report += '\nHistograms:\n'
for key in ['length', 'GC%', 'av.cov', 'mx.cov.', 'av.qual']:
report += '\n'
report += ascii_histogram(numpy.histogram(contigStats[key], bins=bins),
label=key,
**kwargs)
return report