Python/Numpy/Scipy/Matplotlib statistical summary and histogram column of numbers

stats.py

#!/usr/bin/env python
import sys
import numpy
import scipy
import scipy.stats
from optparse import OptionParser

parser = OptionParser()
parser.add_option("-b", "--bins", dest="bins", default=50,
                help="Number of bins (default = 50)")
parser.add_option("-c", "--csv", dest="csv", default=False, action="store_true",
                help="Set flag for csv output.")
(options, args) = parser.parse_args()

data = []
for i in sys.stdin:
    try: 
        data.append(float(i))
    except ValueError: 
        sys.stderr.write("Input not a number, skipping (%s)\n"%str(i.strip()))

if len(data) <= 2: 
    sys.stderr.write("Too few data points. Exiting.\n")
    sys.exit()

res = ( len(data),
        numpy.sum(data),
        numpy.mean(data),
        numpy.median(data),
        numpy.std(data),
        numpy.nanmin(data),
        numpy.nanmax(data),
        scipy.stats.skew(data)   )

if options.csv:
    print ','.join(map(str,res))
else:
    print "n: ....... %d\nsum: ..... %f\nmean: .... %f\nmedian: .. %f\nstd: ..... %f\nmin: ..... %f\nmax: ..... %f\nskew: .... %f"%res

try:    
    import matplotlib
    matplotlib.use('Agg')
    import matplotlib.pyplot as plt

    try:
        nbins = int(options.bins)
    except ValueError:
        nbins = 50
    
    fig = plt.figure()
    plt.hist(data, bins=nbins)
    fig.savefig("hist.png")
    
    # cdf
    fig1 = plt.figure()
    [y, x0, bs, xtra] = scipy.stats.cumfreq(data, nbins) # bin values, lowerlimit, binsize, extrapoints 
    scale = 1.0/y[-1]
    y *= scale
    x = numpy.linspace(x0, x0 + nbins*bs, nbins)
    plt.plot(x, y)
    fig1.savefig("cdf.png")
except ImportError: 
    pass

This is a straightforward trick to generate the histogram. If you like to see the simplest thing, just visit the article where all the things explained very well with code