// the find
Visualize-ML/Linear-Algebra-Made-Easy---Learn-with-Python-and-Visualization
”数学不难“ 之 《线性代数不难》上下册，66话题完册；欢迎批评指正
A bilingual (Chinese/English) linear algebra course delivered as 66 numbered topics, each pairing a PDF slide deck with a Jupyter notebook that visualizes the concept in matplotlib. It's aimed at self-learners and ML practitioners who want geometric intuition for vectors, matrices, eigen/SVD decompositions, and PCA rather than a dry proof-based textbook.
The topic breadth is real — it goes from 'what is a vector' all the way through QR decomposition, spectral clustering, and four flavors of SVD, which is more than most intro courses attempt. Every concept gets a matplotlib visualization directly tied to the math (rotation, shearing, orthogonal projection as actual plotted transforms), not just static diagrams. Full English translations exist for the entire book, not just cherry-picked sections, so it's not a half-translated afterthought.
There's no visible README or requirements.txt in the tree, so a newcomer has to guess whether to start from a PDF or a notebook, and has no pinned environment to run 100+ notebooks reliably. With one or two notebooks per topic and no shared plotting module visible, the matplotlib boilerplate is almost certainly duplicated across the whole repo rather than factored into a small library. The repo is asset-heavy (dozens of multi-MB PDFs plus pkl data files checked into git), which makes it a slow, awkward clone for something that's really a curriculum, not a codebase, and there's no CI or tests since none apply to static teaching material.