Quick and Efficient ways to calculate Pairwise Distances in Python
Introduction
Many a time in machine learning and data analysis projects we work on datasets where distance computation between the instances of data is one of the crucial components. For example, while working on a route optimization problem, before optimizing over the distances of two different locations I wanted to calculate a distance matrix which comprises of pairwise distances between each location.
This is why I wondered what would be the most efficient method to calculate this pairwise distance matrix, which contains the distances between each coordinate when working on a large dataset. In this article, I discuss and compare three such methods to efficiently calculate pair-wise distances between two arrays.
Distance Metrics
While calculating the distances between two given arrays several metrics could be considered as per the requirement of the problem.
Euclidean Distance
This is probably the most common distance metric used in geometry. It measures the (shortest distance) straight line distance between two points (vectors). It is calculated by square rooting the sum of squared differences of the elements of two vectors. It can also be defined as the L2 norm of the…
