def _small_file_innerdist(start,
fastq_file,
pair_file,
ref_file,
out_base,
out_dir,
config,
remove_workdir=False):
work_dir = os.path.join(out_dir, "innerdist_estimate")
if os.path.exists(work_dir):
shutil.rmtree(work_dir)
safe_makedir(work_dir)
extra_args = ["-s", str(start), "-u", "250000"]
bowtie_runner = _select_bowtie_version(config)
out_sam = bowtie_runner.align(fastq_file, pair_file, ref_file, out_base,
work_dir, config, extra_args)
dists = []
with closing(pysam.Samfile(out_sam)) as work_sam:
for read in work_sam:
if read.is_proper_pair and read.is_read1:
dists.append(abs(read.isize) - 2 * read.rlen)
if dists:
dist_stats = Stats(dists)
return int(round(dist_stats.mean())), int(
round(dist_stats.standard_deviation()))
else:
return None, None
def _estimate_paired_innerdist(fastq_file, pair_file, ref_file, out_base,
out_dir, config):
"""Use Bowtie to estimate the inner distance of paired reads.
"""
# skip initial reads for large file, but not for smaller
dists = _bowtie_for_innerdist("1000000", fastq_file, pair_file, ref_file,
out_base, out_dir, config)
if len(dists) == 0:
dists = _bowtie_for_innerdist("1", fastq_file, pair_file, ref_file,
out_base, out_dir, config, True)
dist_stats = Stats(dists)
return int(round(dist_stats.mean())), int(round(dist_stats.standard_deviation()))
def _estimate_paired_innerdist(fastq_file, pair_file, ref_file, out_base,
out_dir, config):
"""Use Bowtie to estimate the inner distance of paired reads.
"""
# skip initial reads for large file, but not for smaller
dists = _bowtie_for_innerdist("1000000", fastq_file, pair_file, ref_file,
out_base, out_dir, config)
if len(dists) == 0:
dists = _bowtie_for_innerdist("1", fastq_file, pair_file, ref_file,
out_base, out_dir, config, True)
dist_stats = Stats(dists)
return int(round(dist_stats.mean())), int(
round(dist_stats.standard_deviation()))
def _bowtie_for_innerdist(start, fastq_file, pair_file, ref_file, out_base,
out_dir, data, remove_workdir=False):
work_dir = os.path.join(out_dir, "innerdist_estimate")
if os.path.exists(work_dir):
shutil.rmtree(work_dir)
safe_makedir(work_dir)
extra_args = ["-s", str(start), "-u", "250000"]
ref_file, bowtie_runner = _determine_aligner_and_reference(ref_file, data["config"])
out_sam = bowtie_runner.align(fastq_file, pair_file, ref_file, {"lane": out_base},
work_dir, data, extra_args)
dists = []
with closing(pysam.Samfile(out_sam)) as work_sam:
for read in work_sam:
if read.is_proper_pair and read.is_read1:
dists.append(abs(read.isize) - 2 * read.rlen)
if dists:
dist_stats = Stats(dists)
return int(round(dist_stats.mean())), int(round(dist_stats.standard_deviation()))
else:
return None, None
def descriptive_stats(xs):
if len(xs) < 2:
return xs
calc = Stats(xs)
parts = ["min: %s" % min(xs),
"5%%: %s" % calc.percentile(5),
"25%%: %s" % calc.percentile(25),
"median: %s" % calc.percentile(50),
"75%%: %s" % calc.percentile(75),
"95%%: %s" % calc.percentile(95),
"99%%: %s" % calc.percentile(99),
"max: %s" % max(xs)]
return "\n".join([" " + x for x in parts])
def descriptive_stats(xs):
if len(xs) < 2:
return xs
calc = Stats(xs)
parts = [
"min: %s" % min(xs),
"5%%: %s" % calc.percentile(5),
"25%%: %s" % calc.percentile(25),
"median: %s" % calc.percentile(50),
"75%%: %s" % calc.percentile(75),
"95%%: %s" % calc.percentile(95),
"99%%: %s" % calc.percentile(99),
"max: %s" % max(xs)
]
return "\n".join([" " + x for x in parts])
class TestPyDescriptiveStatistics(unittest.TestCase):
def setUp(self):
self.enum = Enum([2, 6, 9, 3, 5, 1, 8, 3, 6, 9, 2])
def test_number(self):
self.assertEqual(self.enum.number(), 11)
def test_sum(self):
self.assertEqual(self.enum.sum(), 54)
def test_mean(self):
self.assertEqual(self.enum.mean(), 4.909090909090909)
def test_median(self):
self.assertEqual(self.enum.median(), 5.0)
def test_variance(self):
self.assertEqual(self.enum.variance(), 7.7190082644628095)
def test_standard_deviation(self):
self.assertEqual(self.enum.standard_deviation(), 2.778310325442932)
def test_percentile(self):
self.assertEqual(self.enum.percentile(70), 6.0)
class TestPyDescriptiveStatistics(unittest.TestCase):
def setUp(self):
self.enum = Enum([2,6,9,3,5,1,8,3,6,9,2])
def test_number(self):
self.assertEqual(self.enum.number(), 11)
def test_sum(self):
self.assertEqual(self.enum.sum(), 54)
def test_mean(self):
self.assertEqual(self.enum.mean(), 4.909090909090909)
def test_median(self):
self.assertEqual(self.enum.median(), 5.0)
def test_variance(self):
self.assertEqual(self.enum.variance(), 7.7190082644628095)
def test_standard_deviation(self):
self.assertEqual(self.enum.standard_deviation(), 2.778310325442932)
def test_percentile(self):
self.assertEqual(self.enum.percentile(70), 6.0)
#! /usr/bin/python
import subprocess
from py_descriptive_statistics import Enum # calculates variance using https://github.com/gleicon/py_descriptive_statistics
city = 'cincinnati'
last = subprocess.check_output(["head", "-n", "1", "{}pair.txt".format(city)])
now = subprocess.check_output(["tail", "-n", "1", "{}pair.txt".format(city)])
last = float(last)
now = float(now)
enum = Enum([last, now])
rounded = round(enum.variance())
print rounded
import sys
sys.path.append("..")
from py_descriptive_statistics import Enum
enum = Enum([2,6,9,3,5,1,8,3,6,9,2])
print enum.number()
print enum.sum()
print enum.mean()
print enum.median()
print enum.variance()
print enum.standard_deviation()
print enum.percentile(70)
print enum.percentile(95)
print enum.percentile(99)
def setUp(self):
self.enum = Enum([2, 6, 9, 3, 5, 1, 8, 3, 6, 9, 2])
#! /usr/bin/python
import subprocess
from py_descriptive_statistics import Enum # calculates variance using https://github.com/gleicon/py_descriptive_statistics
city = 'cincinnati'
last = subprocess.check_output(["head", "-n", "1", "{}pair.txt".format(city)])
now = subprocess.check_output(["tail", "-n", "1", "{}pair.txt".format(city)])
last = float(last)
now = float(now)
enum = Enum([last,now])
rounded = round(enum.variance())
print rounded
import sys
sys.path.append("..")
from py_descriptive_statistics import Enum
enum = Enum([2, 6, 9, 3, 5, 1, 8, 3, 6, 9, 2])
print enum.number()
print enum.sum()
print enum.mean()
print enum.median()
print enum.variance()
print enum.standard_deviation()
print enum.percentile(70)
print enum.percentile(95)
print enum.percentile(99)
#! /usr/bin/python
import subprocess
from py_descriptive_statistics import Enum # calculates standard deviation (root of variance?) using https://github.com/gleicon/py_descriptive_statistics
city = 'cincinnati'
last = subprocess.check_output(["head", "-n", "1", "{}pair.txt".format(city)])
now = subprocess.check_output(["tail", "-n", "1", "{}pair.txt".format(city)])
last = float(last)
now = float(now)
enum = Enum([last, now])
rounded = round(enum.standard_deviation()) # rounds the number cleanly
print rounded
def setUp(self):
self.enum = Enum([2,6,9,3,5,1,8,3,6,9,2])
#! /usr/bin/python
import subprocess
from py_descriptive_statistics import Enum # calculates standard deviation (root of variance?) using https://github.com/gleicon/py_descriptive_statistics
city = 'cincinnati'
last = subprocess.check_output(["head", "-n", "1", "{}pair.txt".format(city)])
now = subprocess.check_output(["tail", "-n", "1", "{}pair.txt".format(city)])
last = float(last)
now = float(now)
enum = Enum([last,now])
rounded = round(enum.standard_deviation()) # rounds the number cleanly
print rounded
