Question

I have two lists (which may or may not be the same length). In each list, are a series of tuples of two points (basically X, Y values).

I am comparing the two lists against each other to find two points with similar point values. I have tried list comprehension techniques, but it got really confusing with the nested tuples inside of the lists and I couldn’t get it to work.

Is this the best (fastest) way of doing this? I feel like there might be a more Pythonic way of doing this.

Say I have two lists:

pointPairA = [(2,1), (4,8)]
pointPairB = [(3,2), (10,2), (4,2)]

And then an empty list for storing the pairs and a tolerance value to store only matched pairs

matchedPairs = []
tolerance = 2

And then this loop that unpacks the tuples, compares the difference, and adds them to the matchedPairs list to indicate a match.

for pointPairA in pointPairListA:
    for pointPairB in pointPairListB:
        ## Assign the current X,Y values for each pair
        pointPairA_x, pointPairA_y = pointPairA
        pointPairB_x, pointPairB_x = pointPairB

        ## Get the difference of each set of points
        xDiff = abs(pointPairA_x - pointPairB_x)
        yDiff = abs(pointPairA1_y - pointPairB_y)

        if xDiff < tolerance and yDiff < tolerance:
            matchedPairs.append((pointPairA, pointPairB))

That would result in matchedPairs looking like this, with tuples of both point tuples inside:

[( (2,1), (3,2) ), ( (2,1), (4,2) )]

Answer 1

Here pointpairA is the single list and pointpairB would be one of the list of 20k

from collections import defaultdict
from itertools import product

pointPairA = [(2,1), (4,8)]
pointPairB = [(3,2), (10,2), (4,2)]
tolerance = 2

dA = defaultdict(list)
tolrange = range(-tolerance, tolerance+1)
for pA, dx, dy in product(pointPairA, tolrange, tolrange):
    dA[pA[0]+dx,pA[1]+dy].append(pA)

# you would have a loop here though the 20k lists
matchedPairs = [(pA, pB) for pB in pointPairB for pA in dA[pB]]  

print matchedPairs