SciPy Tutorial

SciPy is a Python library for scientific and technical computing. Its name is short for Scientific Python.

SciPy is built on top of NumPy. NumPy gives you fast arrays, and SciPy adds ready-made tools for statistics, finding the best values (optimization), filling gaps in data (interpolation), calculus and much more.

Why Learn SciPy?

• It saves you from writing difficult maths code yourself; the tested functions are already there
• It is used by engineers, scientists, data analysts and machine learning libraries
• Its statistics module is great for real questions, like "is the new teaching method really better?"
• It works smoothly with NumPy and Pandas

A First Example

Find the lowest point of the curve y = (x - 3)² + 2:

from scipy.optimize import minimize_scalar

result = minimize_scalar(lambda x: (x - 3) ** 2 + 2)
print(round(result.x, 4))
print(round(result.fun, 4))

3.0
2.0

SciPy found that the curve is lowest at x = 3, where y = 2.

What You Will Learn

• Installing and importing SciPy, and its modules
• Scientific constants and unit conversion
• Descriptive statistics and probability distributions
• Hypothesis tests such as the t-test and correlation
• Optimization: finding minimums and roots
• Interpolation and integration

You should know basic Python and the basics of NumPy before you start.

What to Learn After SciPy

Use Pandas to prepare real data for SciPy's tests, or try R, a language designed for statistics.