As you did earlier, you can instantiate and seed the random number generator using np.random.default_rng(42).
An empty array can be initialized using np.empty(), with the number of entries specified inside the parentheses.
Earlier, you looped over a range of 100,000 values and then n values using for i in range(100000) and for i in range(n), respectively. Here, we need to loop over the number of entries we specified. While we have 1000 entries, or simulations, we will only be performing 100 bernoulli trials, each of which has a probability of success p = 0.05; these are the arguments that will need to be passed in to the perform_bernoulli_trials() function that was defined in the previous exercise. For each trial, the result should be assigned to the appropriate index of n_defaults.
You can plot a histogram using plt.hist(), but remember to specify density=True in addition to n_defaults.
Use plt.show() to display the plot.

What is the goal of statistical inference?

Statistical Thinking in Python (Part 1)