NumPy array filtering with boolean masks

Filtering is the first thing you do with real data: keep the rows you need. In NumPy you do it without loops, with a boolean mask. Edit the code and press Run.

Python 3.12 in your browser (Pyodide)Term: Boolean indexing

Step 1: a mask is an array of True and False

Compare an array with a value and you get one True or False per element. Use it inside [ ] to keep only the True ones.

Code for Step 1: a mask is an array of True and False

import numpy as np
marks = np.array([35, 78, 92, 41, 66, 88, 23])
mask = marks >= 40
print(mask)
print(marks[mask])
print("passed:", mask.sum(), "of", marks.size)

Step 2: combine conditions

Use & (and), | (or) and ~ (not), with brackets around each condition.

Code for Step 2: combine conditions

import numpy as np
prices = np.array([499, 1299, 89, 2999, 650, 15000])
mid = prices[(prices > 400) & (prices < 3000)]
print(mid)
print(prices[~(prices > 1000)])

Step 3: np.where and filtering a table

np.where picks values by a condition. Filtering one column decides the rows of another.

Code for Step 3: np.where and filtering a table

import numpy as np
city = np.array(["Pune", "Delhi", "Pune", "Chennai"])
orders = np.array([12, 30, 8, 15])
print(np.where(orders > 10, "busy", "quiet"))
print(orders[city == "Pune"].sum())