Skip to content
Ephizen logo ephizen / wiki
§ cs 403 · 6 guides

CS 403 — NumPy & Pandas

Arrays, broadcasting, vectorization, dataframes, groupby, joins — numeric and tabular work in Python. The substrate every ML training script sits on.

Level
6 guides
# Code Name Level Updated
1 —
NumPy Arrays — dtype, shape, strides What an ndarray actually is in memory — the prerequisite to making any of it fast.
intermediate
2 —
NumPy Broadcasting The shape-alignment rules that let `(N,1) * (1,M)` produce an `(N,M)` matrix without a loop.
intermediate
3 —
Pandas DataFrames — Fundamentals Series, DataFrame, indexing, dtypes — the parts you actually use, plus the parts that bite.
beginner
4 —
Pandas groupby — Split-Apply-Combine Aggregation, transformation, and filtration over groups — feature engineering's workhorse.
intermediate
5 —
Pandas merge / join Inner, outer, left, right, and the merge keys that decide whether your join row-explodes.
intermediate
6 —
Vectorization vs Loops Why a NumPy one-liner runs 100× faster than the Python loop that does the same thing.
intermediate