Answers

Creating a mask containing a circle

Pair programming: part 1

Exercise A

Create a 100 x 100 square grid containing a circle. The centre of the circle should be at position x = 40, y = 50 and have a radius of 10. Call your output circular_mask.

Break this idea down into the different steps: - Start by creating an array - how could you create an array of the right shape? - How would you turn the equation above into code? - How would you evaluate one (x, y) position to start with? - How do you move on to evaluate every position? - …

Important note: For numpy remember that the axes are opposite to what you would expect for an x-y grid - they store data as row-major meaning you access the row as the first index (y) and the column as the second index (x). This means position arr[1, 0] would be at position (0, 1) in x, y co-ordinates rather than (1, 0) as you may expect.

When completed, you can use the plotting code below to show your mask. This uses a matplotlib function called imshow (documentation; example from image gallery) which can plot your 2D data on a grid.

Extended:

Consider how you could create a mask for a rectangle rather than a circle (or a square)

---
title: Answers
jupyter: python3
---


# Creating a mask containing a circle

## Pair programming: part 1


### Exercise A

Create a 100 x 100 square grid containing a circle. The centre of the circle should be at position `x = 40`, `y = 50` and have a radius of 10. Call your output `circular_mask`.

*Break this idea down into the different steps:*
 - *Start by creating an array - how could you create an array of the right shape?*
 - *How would you turn the equation above into code?*
 - *How would you evaluate one (x, y) position to start with?*
 - *How do you move on to evaluate every position?*
 - *...*

**Important note**: For `numpy` remember that the axes are opposite to what you would expect for an x-y grid - they store data as *row-major* meaning you access the row as the first index (y) and the column as the second index (x). This means position `arr[1, 0]` would be at position (0, 1) in x, y co-ordinates rather than (1, 0) as you may expect.

When completed, you can use the plotting code below to show your mask. This uses a `matplotlib` function called `imshow` ([documentation](https://matplotlib.org/stable/api/_as_gen/matplotlib.pyplot.imshow.html); [example from image gallery](https://matplotlib.org/stable/gallery/images_contours_and_fields/image_demo.html)) which can plot your 2D data on a grid.

```{pyodide}
#| caption: "▶ Ctrl/Cmd+Enter | ⇥ Ctrl/Cmd+] | ⇤ Ctrl/Cmd+["
# ADD CODE HERE
import numpy as np

n  = 100
circular_mask = np.zeros((n,n))
x0, y0 = 40,50
radius = 10
rows, cols = np.indices(circular_mask.shape)

circular_mask = ((rows-y0)**2+(cols-x0)**2)<=radius**2

```

```{pyodide}
#| caption: "▶ Ctrl/Cmd+Enter | ⇥ Ctrl/Cmd+] | ⇤ Ctrl/Cmd+["
### UNCOMMENT THIS CODE TO PLOT YOUR MASK WHEN COMPLETED

import matplotlib.pyplot as plt

fig, ax = plt.subplots(figsize=(6, 3.2))

ax.set_title('Circular Mask')
ax.imshow(circular_mask,cmap="bone",origin="lower") # Setting colourmap (cmap) and the position of 0,0 (origin) on the plot
```

---

*Extended*:

1. Consider how you could create a mask for a rectangle rather than a circle (or a square)

```{pyodide}
#| caption: "▶ Ctrl/Cmd+Enter | ⇥ Ctrl/Cmd+] | ⇤ Ctrl/Cmd+["
# 10 by 20 rectangle from x0,y0
Lx = 10
Ly= 20
mask = (rows-y0>0) & (rows-y0<Ly)  & (cols-x0>0) & (cols-x0<Lx)
plt.imshow(mask,cmap="bone",origin="lower")
```