CODEWITHSUNDEEP
pythonnumpyline
user1022585
user1022585

Reputation: 13641

'line of sight' from numpy array in Python

I'm using a numpy array of 0s and 1s to represent collision in a game map, co-ords set as 1 represent a wall.

What I need to do is a line of sight test from point A to B. So, given the values startX, startY, endX and endY, I need to get all values from my numpy array in a straight line, and if any of those are a wall (1) then there is no line of sight.

Is there a way to get a line of values from numpy? Or is this probably the wrong way to go about it?

Any pointers?

Upvotes: 1

Views: 2500

Answers (2)

AfroDisco
AfroDisco

Reputation: 41

As said Daniel F, this can be an implementation. I improved, completed and fixed its implementation.

import numpy as np

a = np.ones([7, 7])
a[3, 3] = 1
a[1, :] = 2
a[4, :] = 2

b = 1 - a
b = b.astype(bool)

startX = 3
startY = 0
endX = 3
endY = 5
n = 4  # Steps per unit distance
dxy = (np.sqrt((endX - startX) ** 2 + (endY - startY) ** 2)) * n
i = np.rint(np.linspace(startX, endX, dxy)).astype(int)
j = np.rint(np.linspace(startY, endY, dxy)).astype(int)
has_collision = np.any(b[i, j])

print("Has collision", has_collision)

Upvotes: 2

Daniel F
Daniel F

Reputation: 14399

A simple implementation

coll = np.random.randint(0, 2, (100,100)) # collision matrix
n = 4 # bigger for more accurate, less speed
dxy = (startX - endX + startY - endY) * n
i = np.rint(np.linspace(startX, endX, dxy))
j = np.rint(np.linspace(startY, endY, dxy))
collision = np.any(coll[i, j])

Upvotes: -1

Related Questions