Mastering pandas - Second Edition

Uniform distribution

Suppose we have a probability distribution of X that means the following is true:

Pr(X=x) = 1/b-a for all a<X<b

Pr(X=x) = 0 for all other X

Here, a and b are parameters of uniform distribution.

Since the probability is the same (or uniform) for all the values, it is called a uniform distribution.

Suppose there are n data points in a dataset that is hypothesized to follow a uniform distribution. Based on these points, we aim to find the values of a and b to define the distribution to which these data points most likely belong. For this, we can use the maximum likelihood estimate method:

We have to find the a, b that maximises ...