PyTorch 1.x Reinforcement Learning Cookbook

How to do it...

Let’s explore the beta distribution through the following steps:

1. Import PyTorch and matplotlib because we will visualize the shape of the distributions:

>>> import torch>>> import matplotlib.pyplot as plt

2. We start by visualizing the shape of the beta distribution with the starting positions, α=1 and β=1:

>>> beta1 = torch.distributions.beta.Beta(1, 1)>>> samples1 = [beta1.sample() for _ in range(100000)]>>> plt.hist(samples1, range=[0, 1], bins=10)>>> plt.title(‘beta(1, 1)’)>>> plt.show()

You will see the following plot:

Obviously, when α=1 and β=1, it doesn't provide any information about where the true value lies in the ...