> For the complete documentation index, see [llms.txt](https://python-self-driving.gitbook.io/self-driving/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://python-self-driving.gitbook.io/self-driving/python/why-use-numpy-arrays.md).

# 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])`

```python
[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.

```python
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))

```

```python
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)`