Hichem MG

Posted on Apr 23

📐 NumPy Math Functions Cheat Sheet

#python #tutorial #math #numpy

NumPy is Python’s go-to library for numerical computing, especially with large datasets and arrays. Its math functions are vectorized, which means you can apply operations to entire arrays without loops — making them fast and powerful.

Why Use NumPy for Math?

Element-wise operations on arrays
Handles multi-dimensional data
Functions are vectorized (fast & efficient)
Seamless with machine learning, data analysis, and scientific computing

📦 Importing NumPy

import numpy as np

Note:

The following are the most-used functions only, checkout the complete list on numpy.org.

🔢 Arithmetic Functions

Function	Description
`np.add(x1, x2)`	Element-wise addition
`np.subtract(x1, x2)`	Element-wise subtraction
`np.multiply(x1, x2)`	Element-wise multiplication
`np.divide(x1, x2)`	Element-wise division
`np.power(x1, x2)`	x1 raised to the power of x2
`np.mod(x1, x2)`	Element-wise modulo

📐 Trigonometric Functions

Function	Description
`np.sin(x)`	Sine
`np.cos(x)`	Cosine
`np.tan(x)`	Tangent
`np.arcsin(x)`	Inverse sine
`np.arccos(x)`	Inverse cosine
`np.arctan(x)`	Inverse tangent
`np.arctan2(y, x)`	Arctangent of y/x
`np.hypot(x, y)`	√(x² + y²)
`np.degrees(x)`	Radians → degrees
`np.radians(x)`	Degrees → radians

📊 Exponentials & Logarithms

Function	Description
`np.exp(x)`	e ** x
`np.expm1(x)`	e ** x - 1
`np.log(x)`	Natural log (base e)
`np.log2(x)`	Base-2 logarithm
`np.log10(x)`	Base-10 logarithm
`np.log1p(x)`	log(1 + x) (accurate for small x)

🧠 Rounding, Absolute, and Sign Functions

Function	Description
`np.round(x)`	Round to nearest integer
`np.floor(x)`	Round down
`np.ceil(x)`	Round up
`np.trunc(x)`	Truncate decimal
`np.abs(x)`	Absolute value
`np.sign(x)`	Sign of x (-1, 0, or 1)

📐 Aggregates & Reductions

Function	Description
`np.sum(x)`	Sum of all elements
`np.prod(x)`	Product of all elements
`np.mean(x)`	Average value
`np.std(x)`	Standard deviation
`np.var(x)`	Variance
`np.max(x)`	Maximum value
`np.min(x)`	Minimum value

🔍 Comparisons & Logical Checks

Function	Description
`np.isfinite(x)`	Finite check
`np.isinf(x)`	Infinite check
`np.isnan(x)`	NaN check
`np.isclose(x1, x2)`	Check if values are close
`np.equal(x1, x2)`	Element-wise equality

📏 Linear Algebra (Bonus)

Function	Description
`np.dot(x1, x2)`	Dot product
`np.matmul(x1, x2)`	Matrix multiplication
`np.linalg.inv(x)`	Inverse of a matrix
`np.linalg.det(x)`	Determinant
`np.linalg.eig(x)`	Eigenvalues and eigenvectors

💡 Constants

Constant	Description
`np.pi`	π
`np.e`	Euler’s number
`np.inf`	Infinity
`np.nan`	Not a number

⚙️ Practical NumPy Example

You can run it and see the output online here: pythononline.net/#qBOCl4

import numpy as np

arr = np.array([1, 2, 3, 4])
arr2 = np.array([4, 3, 2, 1])

# Arithmetic
print("Add:", np.add(arr, arr2))
print("Subtract:", np.subtract(arr, arr2))
print("Multiply:", np.multiply(arr, arr2))
print("Divide:", np.divide(arr, arr2))
print("Power:", np.power(arr, 2))
print("Mod:", np.mod(arr, 3))

# Trigonometry
print("sin(pi/2):", np.sin(np.pi / 2))
print("cos(0):", np.cos(0))
print("tan(pi/4):", np.tan(np.pi / 4))
print("arcsin(1):", np.arcsin(1))
print("arccos(0):", np.arccos(0))
print("arctan(1):", np.arctan(1))
print("arctan2(1, 1):", np.arctan2(1, 1))
print("hypot(3, 4):", np.hypot(3, 4))
print("degrees(pi):", np.degrees(np.pi))
print("radians(180):", np.radians(180))

# Exponentials & logs
print("exp(1):", np.exp(1))
print("expm1(1e-5):", np.expm1(1e-5))
print("log(e):", np.log(np.e))
print("log2(8):", np.log2(8))
print("log10(1000):", np.log10(1000))
print("log1p(1e-5):", np.log1p(1e-5))

# Absolute, rounding
print("abs(-5):", np.abs(-5))
print("sign(-3):", np.sign(-3))
print("round(3.1415):", np.round(3.1415))
print("floor(3.9):", np.floor(3.9))
print("ceil(3.1):", np.ceil(3.1))
print("trunc(3.8):", np.trunc(3.8))

# Aggregates
print("sum:", np.sum(arr))
print("prod:", np.prod(arr))
print("mean:", np.mean(arr))
print("std:", np.std(arr))
print("var:", np.var(arr))
print("max:", np.max(arr))
print("min:", np.min(arr))

# Comparisons
print("isfinite([1, np.inf]):", np.isfinite([1, np.inf]))
print("isinf([1, np.inf]):", np.isinf([1, np.inf]))
print("isnan([1, np.nan]):", np.isnan([1, np.nan]))
print("isclose(1.0, 1.00001):", np.isclose(1.0, 1.00001))
print("equal([1, 2], [1, 3]):", np.equal([1, 2], [1, 3]))

# Constants
print("pi:", np.pi)
print("e:", np.e)
print("inf:", np.inf)
print("nan:", np.nan)

NumPy offers a rich set of fast, vectorized math functions that go far beyond Python’s built-ins. Whether you're handling basic arithmetic, advanced trigonometry, or high-performance numerical computations, NumPy is an essential tool.

Keep this cheat sheet handy as a quick reference to power up your data science, machine learning, or scientific computing projects!

Top comments (1)

Keithwalker • Aug 26

This cheat sheet is super helpful! NumPy really shines when working with large datasets since its math functions are vectorized and lightning fast compared to plain Python loops. Whether it’s arithmetic, trigonometry, or linear algebra, you can handle it all efficiently with Python arrays. Having these functions at your fingertips saves time and makes code cleaner, especially for data science, ML, or web projects involving heavy computations.

CodeNewbie Community 🌱