python plot and powerlaw fit

Question

I have the following list:

[6, 4, 0, 0, 0, 0, 0, 1, 3, 1, 0, 3, 3, 0, 0, 0, 0, 1, 1, 0, 0, 0, 3, 2, 3, 3, 2, 5, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 2, 1, 0, 1, 0, 0, 0, 0, 1, 0, 1, 2, 0, 0, 0, 2, 1, 1, 0, 0, 0, 1, 0, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 3, 1, 1, 0, 1, 1, 0, 0, 0, 1, 0, 0, 1, 0, 1, 2, 2, 3, 2, 1, 0, 0, 0, 1, 2]

I want to plot the frequency of each entity with python and make a powerlaw analysis on it.

But I cannot figure how I can plot the list with ylabel the frequency and xlabel the numbers on the list.

I thought to create a dict with the frequencies and plot the values of the dictionary, but with that way, I cannot put the numbers on xlabel.

Any advice?

Filariasis

Answer 1

I think you're right about the dictionary:

>>> import matplotlib.pyplot as plt
>>> from collections import Counter
>>> c = Counter([6, 4, 0, 0, 0, 0, 0, 1, 3, 1, 0, 3, 3, 0, 0, 0, 0, 1, 1, 0, 0, 0, 3, 2, 3, 3, 2, 5, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 2, 1, 0, 1, 0, 0, 0, 0, 1, 0, 1, 2, 0, 0, 0, 2, 1, 1, 0, 0, 0, 1, 0, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 3, 1, 1, 0, 1, 1, 0, 0, 0, 1, 0, 0, 1, 0, 1, 2, 2, 3, 2, 1, 0, 0, 0, 1, 2])
>>> sorted(c.items())
[(0, 50), (1, 30), (2, 9), (3, 8), (4, 1), (5, 1), (6, 1)]
>>> plt.plot(*zip(*sorted(c.items()))
... )
[<matplotlib.lines.Line2D object at 0x36a9990>]
>>> plt.show()

There are a few pieces here that are of interest. zip(*sorted(c.items())) will return something like [(0,1,2,3,4,5,6),(50,30,9,8,1,1,1)]. We can unpack that using the * operator so that plt.plot sees 2 arguments -- (0,1,2,3,4,5,6) and (50,30,9,8,1,1,1). which are used as the x and y values in plotting respectively.

As for fitting the data, scipy will probably be of some help here. Specifically, have a look at the following examples. (one of the examples even uses a power law).

Verse

I just see your edits. Thank you. This will probably solve my problem. – Filariasis

Answer 2

Use the package: powerlaw

import powerlaw
d=[6, 4, 0, 0, 0, 0, 0, 1, 3, 1, 0, 3, 3, 0, 0, 0, 0, 1, 1, 0, 0, 0, 3,2,  3, 3, 2, 5, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 2, 1, 0, 1, 0, 0, 0, 0, 1,0, 1, 2, 0, 0, 0, 2, 1, 1, 0, 0, 0, 1, 0, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1,3, 1, 1, 0, 1, 1, 0, 0, 0, 1, 0, 0, 1, 0, 1, 2, 2, 3, 2, 1, 0, 0, 0, 1, 2]
fit = powerlaw.Fit(numpy.array(d)+1,xmin=1,discrete=True)
fit.power_law.plot_pdf( color= 'b',linestyle='--',label='fit ccdf')
fit.plot_pdf( color= 'b')

print('alpha= ',fit.power_law.alpha,'  sigma= ',fit.power_law.sigma)

alpha= 1.85885487521 sigma= 0.0858854875209

It allow to plot, fit and analyse the data correctly. It has as special method for fit on power law distributions with discrete data.

it can be installed with: pip install powerlaw

Caprifig

Do you know by any chance how to get the scaling factor C? – Heikeheil

Answer 3

y = np.bincount([6, 4, 0, 0, 0, 0, 0, 1, 3, 1, 0, 3, 3, 0, 0, 0, 0, 1, 1, 0, 0, 0, 3, 2, 3, 3, 2, 5, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 2, 1, 0, 1, 0, 0, 0, 0, 1, 0, 1, 2, 0, 0, 0, 2, 1, 1, 0, 0, 0, 1, 0, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 3, 1, 1, 0, 1, 1, 0, 0, 0, 1, 0, 0, 1, 0, 1, 2, 2, 3, 2, 1, 0, 0, 0, 1, 2])
x = np.nonzero(y)[0]
plt.bar(x,y)

Apocope

Answer 4

import matplotlib.pyplot as plt
data = [6, 4, 0, 0, 0, 0, 0, 1, 3, 1, 0, 3, 3, 0, 0, 0, 0, 1, 1, 0, 0, 0, 3, 2, 3, 3, 2, 5, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 2, 1, 0, 1, 0, 0, 0, 0, 1, 0, 1, 2, 0, 0, 0, 2, 1, 1, 0, 0, 0, 1, 0, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 3, 1, 1, 0, 1, 1, 0, 0, 0, 1, 0, 0, 1, 0, 1, 2, 2, 3, 2, 1, 0, 0, 0, 1, 2]

plt.hist(data, bins=range(max(data)+2))
plt.show()

Surefire