2

I am trying to fit a power law to a histogram (more exact Pareto distribution). I did it with my own function, where I check for smallest sum of squares of difference. But this means I need to loop threw all the coefitients, which can take some time. Another problem is that I need to make my own data list so that I have histogram data.

So I am looking for a function that would return a list of data made by matplotlib.pyplot.hist() and not just a picture and than I would like to fit this data with pareto distribution abit faster than looping so many times and obtain the coefitions.

NPE
  • 486,780
  • 108
  • 951
  • 1,012
user1035617
  • 311
  • 2
  • 5
  • 9
  • Duplicate of http://stackoverflow.com/questions/3242326/fitting-a-pareto-distribution-with-python-scipy – keflavich Jan 12 '12 at 00:12

1 Answers1

3

I think you are looking for the values and the bin sizes.

The matplotlib.pyplot.hist() function returns a tupe with (n, bins, patches)

For more information about this function click this link

For example to plot some 'data', 150 bins:

import matplotlib.pyplot as plt
hist = plt.hist(data,150)
binsize = hist[0]
value = hist[1]
print binsize
print ''
print value
Jon Lin
  • 142,182
  • 29
  • 220
  • 220
Loran de Vries
  • 31
  • 2