Generating Random Numbers in NumPy

Set 1 - Testing Flatness randint

Use np.zeros to create an array with 10 elements, name it counts.
Use a for loop to create 100,000 np.random.randint numbers less than 10. Add 1.0 to each element of counts that your random integer matches. (For instance, if your random number is 0, you would do counts[0] += 1.0. If your random number is 8, you would do counts[8] += 1.0.)
Create a new array bins where you divide each element in counts by 100,000.
Are all 10 spots in bins the same? If np.random.randint truly generated a "flat" random distribution, they should be. Perhaps they would look closer to 10% each if you did a bigger test. Try parts 1 through 3 again with more than 100,000 trials.

Use np.zeros to create an array with 10 elements, name it counts.
Use a for loop to create 100,000 random decimals between zero and one (use rand). For each number you generate, multiple it by 10 and convert it to an integer using int(). Then add your number to the counts bin, as we did in part 2 of set 2.
Create a new array bins where you divide each element in counts by 100,000.
Are all 10 spots in bins the same? If np.random.rand truly generated a "flat" random distribution, they should be. Perhaps they would look closer to 10% each if you did a bigger test. Try parts 1 through 3 again with more than trials.

Unlike randint and rand in Set 1, randn is not supposed to be flat. It is supposed to be a Normal Distribution.

Use zeros to create an array with 10 elements, name it counts.
Use a for loop to create 100,000 random decimals using randn. Take the absolute value (abs) of each of your numbers, then convert it to an integer using int. If your number is less than 10, add 1.0 to the appropriate element in your counts array, as in Sets 1 and 2 above.
Create a new array bins where you divide each element in counts by 100,000.
Take a look at bins, does it match a Normal Distribution? Perhaps it would look closer to Normal if you did a bigger test. Try parts 1 through 3 again with more trials.