def CountDistribution(self):
"""print distribution of number of units and sequences per
family.
"""
self.PrintStatus()
print """#
# NUM: number of units/sequences per family
# NUNITS: number of families with x units
# NSEQ: number of families with x sequences
#"""
print "NUM\tNUNITS\tNSEQ"
sys.stdout.flush()
histograms = []
class_name = self.mTableDomains.GetFieldNameClass()
statement = "SELECT COUNT(*) FROM %s GROUP BY %s" % (
self.mTableNameDomains, class_name)
data = map(lambda x: x[0], self.dbhandle.Execute(statement).fetchall())
h1 = Histogram.Calculate(data)
histograms.append(h1)
statement = "SELECT COUNT(DISTINCT nid) FROM %s GROUP BY %s" % (
self.mTableNameDomains, class_name)
data = map(lambda x: x[0], self.dbhandle.Execute(statement).fetchall())
h2 = Histogram.Calculate(data)
histograms.append(h2)
ch = Histogram.Combine(histograms)
Histogram.Print(ch)
def xCountDistribution( self ):
"""print distribution of number of units and sequences per
family.
"""
self.PrintStatus()
print """#
# NUM: number of units/sequences per family
# NUNITS: number of families with x units
# NSEQ: number of families with x sequences
#"""
print "num\tnunits\tnseq"
sys.stdout.flush()
histograms = []
statement = "SELECT nunits, COUNT(*) FROM %s GROUP BY nunits" % self.mTableNameFamilies
h1 = self.dbhandle.Execute( statement ).fetchall()
histograms.append( h1 )
statement = "SELECT nsequences, COUNT(*) FROM %s GROUP BY nsequences" % self.mTableNameFamilies
h2 = self.dbhandle.Execute( statement ).fetchall()
histograms.append( h2 )
ch = Histogram.Combine( histograms )
Histogram.Print(ch)
def DomainDistribution( self ):
"""
distribution of domains per sequence
"""
self.PrintStatus()
if self.mLogLevel >= 1:
sys.stdout.flush()
print """#
# COUNTS: number
# NDOMAINS: number of domains per sequence
# NDOMAINS/SIN: number of domains without singletons per sequence
# NMOBILES: number of mobile modules per sequence
#"""
print "length\tndomains\tndomains/sin\tnmobiles"
sys.stdout.flush()
histograms = []
statement = "SELECT COUNT(*) FROM %s GROUP BY nid" % self.mTableNameDomains
d1 = map(lambda x: x[0], self.dbhandle.Execute( statement ).fetchall())
histograms.append( Histogram.Calculate( d1 ) )
statement = "SELECT COUNT(*) FROM %s AS d, %s AS f WHERE f.family = d.family AND f.nunits > 1 GROUP BY d.nid" % (self.mTableNameDomains, self.mTableNameFamilies)
d2 = map(lambda x: x[0], self.dbhandle.Execute( statement ).fetchall())
histograms.append( Histogram.Calculate( d2 ) )
statement = "SELECT COUNT(*) FROM %s AS d, %s AS f WHERE f.family = d.family GROUP BY d.nid" % (self.mTableNameDomains, self.mTableNameSubset)
d3 = map(lambda x: x[0], self.dbhandle.Execute( statement ).fetchall())
histograms.append( Histogram.Calculate( d3 ) )
ch = Histogram.Combine( histograms )
Histogram.Print(ch)
def LengthDistribution( self ):
"""print distribution of unit length of families
"""
self.PrintStatus()
print """#
# LENGTH: number of units/sequences per family
# NUNITS: number of domains of that length
# NUNITS/SIN: number of domains without singletons
# NSIN: number of singletons with that length
#"""
print "length\tnunits\tnunits/sin\tnsin"
sys.stdout.flush()
histograms = []
statement = "SELECT CEILING((end-start+1)/10) * 10 AS olength, COUNT(*) FROM %s GROUP BY olength" % self.mTableNameDomains
h1 = self.dbhandle.Execute( statement ).fetchall()
histograms.append( h1 )
statement = "SELECT CEILING((end-start+1)/10) * 10 AS dlength, COUNT(*) FROM %s AS a, %s AS d " %\
(self.mTableNameDomains, self.mTableNameFamilies) +\
" WHERE d.family = a.family AND d.nunits > 1 GROUP BY dlength"
h2 = self.dbhandle.Execute( statement ).fetchall()
histograms.append( h2 )
statement = "SELECT CEILING((end-start+1)/10) * 10 AS alength, COUNT(*) FROM %s AS a, %s AS d " %\
(self.mTableNameDomains, self.mTableNameFamilies) +\
" WHERE d.family = a.family AND d.nunits = 1 GROUP BY alength"
h3 = self.dbhandle.Execute( statement ).fetchall()
histograms.append( h3 )
ch = Histogram.Combine( histograms )
Histogram.Print(ch)
continue
header.append(filename)
infile = open(filename, "r")
h = []
while 1:
line = infile.readline()
if not line: break
if line[0] == "#": continue
if not re.match("(\d+)", line): continue
data = map(string.atof, re.split("\s+", line[:-1]))
h.append((data[0], tuple(data[1:])))
infile.close()
histograms.append(h)
print "# bin\t" + string.join(header, "\t\t")
ch = Histogram.Combine(histograms)
Histogram.Print(ch)
ch = Histogram.Normalize(ch)
print "# bin\t" + string.join(header, "\t\t")
Histogram.Print(ch)
if options.normalize: h = Histogram.Normalize(h)
if options.cumulative: h = Histogram.Cumulate(h)
if options.reverse_cumulative:
h = Histogram.Cumulate(h, direction=0)
hists.append(h)
for m in options.append:
if m == "normalize":
hists.append(Histogram.Normalize(h))
if options.headers:
titles.append(options.headers[x])
elif options.titles:
titles.append(options.titles[x])
else:
titles.append("col%i" % options.columns[x])
if titles:
options.stdout.write("bin\t" + "\t".join(titles) + "\n")
if len(hists) == 1:
Histogram.Print(hists[0], nonull=options.nonull)
else:
combined_histogram = Histogram.Combine(
hists, missing_value=options.missing_value)
Histogram.Print(combined_histogram, nonull=options.nonull)
Experiment.Stop()
