numpy arrays

Numpy is one of those tools that will keep showing up as you learn about self driving cars.

This is because numpy arrays tend to be:

1. compact (they don't take up as much space in memory as a Python list).

2. efficient (computations usually run quicker on numpy arrays then Python lists).

3. convenient (which we'll talk about more now).

consider this 2d python grid (list of lists)

grid = [ [0, 1, 5], [1, 2, 6], [2, 3, 7], [3, 4, 8] ]

It's easy to print, for example, row number 0:

print(grid[0])

[0, 1, 5]

but how would you print COLUMN 0? In numpy, this is easy

import numpy as np

np_grid = np.array([ [0, 1, 5], [1, 2, 6], [2, 3, 7], [3, 4, 8] ])

The ':' usually means "all values

print(np_grid[:,0])

[0 1 2 3]

Using numpy with lists

The numpy library lets you run mathematical expressions on elements of a list. The math library cannot do this. Study the examples below and then run the code cell.

print('\nExample of squaring elements in a list')
print(np.square([1, 2, 3, 4, 5]))

print('\nExample of taking the square root of a list')
print(np.sqrt([1, 4, 9, 16, 25]))

print('\nExamples of taking the cube of a list')
print(np.power([1, 2, 3, 4, 5], 3))

Example of squaring elements in a list
[ 1  4  9 16 25]

Example of taking the square root of a list
[ 1.  2.  3.  4.  5.]

Examples of taking the cube of a list
[  1   8  27  64 125]

What if you wanted to change the shape of the array?

For example, we can turn the 2D grid from above into a 1D array. Here, the -1 means automatically fit all values into this 1D shape

np_1D = np.reshape(np_grid, (1, -1))

print(np_1D)

[[0 1 5 1 2 6 2 3 7 3 4 8]]

We can also create a 2D array of zeros or ones

which is useful for car world creation and analysis. For example, create a 5x4 array

zero_grid = np.zeros((5, 4))

print(zero_grid)