SciPy
Scientific computing algorithms built on top of NumPy.
CurrentadvancedGuide only -- no course yet
Overview
SciPy extends NumPy with algorithms for optimization, statistics, signal processing, and linear algebra -- the tools a scientist or engineer would otherwise implement from scratch, built and tested once as a shared library.
- What it is
- A Python library of scientific-computing algorithms (statistics, optimization, linear algebra) built on NumPy arrays.
- Why it's used
- For statistical tests, optimization problems, and numerical algorithms that would be error-prone to reimplement from scratch.
- Where it fits
- Built on NumPy; used alongside Pandas in scientific and statistical data science work.
Core concepts
- Statistical tests (scipy.stats)
- Optimization (scipy.optimize)
- Linear algebra (scipy.linalg)
- Signal processing
Example
A t-test compares two groups' means -- SciPy provides the tested, correct implementation rather than requiring you to derive the statistics from scratch.
from scipy import stats
result = stats.ttest_ind([23, 25, 22], [30, 32, 29])
print(result.pvalue)Common use cases
- Statistical hypothesis testing
- Numerical optimization
- Scientific and engineering computation
Project ideas
- Run a statistical test comparing two small datasets and interpret the result