- Learn the basics of using
matplotlibto display probability density functions, which we approximate with histograms - Use a list comprehension to collect values
- Slice a list
The end result is file hist.py.
We previously learned how to generate pseudorandom numbers. More specifically, these random numbers follow the uniform distribution. If we ask for 1000 pseudorandom numbers in U(2,8) using our runif(2,8) function, we should see a distribution that looks like the following.
The question is how do we get this display. First, let's figure out how to get 1000 random numbers using our old code:
from runif import *
N = 1000
X = [runif(2,8) for i in range(N)]
print X[0:4] # print elements 0, 1, 2, and 3
That last X[0:4] is called a slice operator. You can do lots of things like: X[5:] elements five through the end, X[0:-1] gives all but the last element.
In order to get a histogram, we need some boilerplate code:
import matplotlib.pyplot as plt
fig = plt.figure() # get a handle on the figure object itself
... plotting stuff ...
plt.show()To plot the histogram, we just use:
plt.hist(X, normed=1)If we want to say that image, we insert this before they show():
plt.savefig('unif-2-8-density.png', format="png")which is what I did to get that image above.
Graphs should always have the axes labeled. Let’s do that as well as add a title and set the range of the graph. Put this code right before the savefig() or show():
plt.title("U(2,8) Density Demo")
plt.xlabel("X", fontsize=16)
plt.ylabel("Density", fontsize=16)
plt.axis([1, 9, 0, .25])
plt.text(2,.22, 'N = %d' % N, fontsize=16)Now set N=30000 and try again. You should get a much more even distribution:
