Distribution plot of an array

问题:

I have a numpy array containing float values in [-10..10]. I would like to plot a distribution-graph of the values, like this (here it is done for a binomial random variable) :

For example I would like bars counting the number of elements in each interval [-10, -9.5], [-9.5, -9], ..., [9.5, 10].

How to prepare such a distribution plot with Python?

回答1:

Indeed matplotlib, more precisely you'll find samples of code corresponding to what you are after at: http://matplotlib.org/examples/pylab_examples/histogram_demo_extended.html

import numpy as np
import matplotlib.pyplot as plt
mu, sigma = 200, 25
x = mu + sigma*np.random.randn(10000)
n, bins, patches = plt.hist(x)
plt.show()

n contains the number of points in each bin and bins the cut off values which are in my example generated automatically. You can of course play with plt.hist's options to obtain the graph that you wish.

In your case, just replace x by your array, and play with the bins option for cut off values e.g.:

plt.hist(x, bins = [-10, -9.5, -9])

You can also simlply pass a scalar n to bins in which case plt.hist will determine cut off values to display a nice graph with n bins.