How harisiqbal88/plotneuralnet works
PlotNeuralNet is a diagram-authoring toolkit: users describe neural-network layouts in Python or LaTeX, then use reusable layer and block definitions to produce TikZ/LaTeX diagrams. The Python path builds snippets and writes a .tex file; the repository also provides hand-authored LaTeX examples. The README describes running a shell script with a TeX installation, but the script’s behavior was not sampled, so compilation details are not expanded here.
Main entry points: tikzeng.py TikZ primitives · unet.py Python architecture · blocks.py Reusable blocks · examples LaTeX examples
Components 9
Diagram authoring
- Python architecture
pyexamples/unet.py - LaTeX examples
examples
Python interface
- Reusable blocks
pycore/blocks.py - TikZ primitives
pycore/tikzeng.py
LaTeX rendering
- Generated TeX
- Layer styles
layers/init.tex - TeX renderer
- Diagram output
Other components
- Diagram author
Connections 9
- Diagram author → Python architecture: authors layout
README.md - Diagram author → LaTeX examples: uses examples
README.md - Python architecture → Reusable blocks: uses blocks
pyexamples/unet.py - Python architecture → TikZ primitives: builds with
pyexamples/unet.py - Reusable blocks → TikZ primitives: delegates to
pycore/blocks.py - TikZ primitives → Generated TeX: writes TeX
pycore/tikzeng.py - TikZ primitives → Layer styles: includes styles
pycore/tikzeng.py - Generated TeX → TeX renderer: compiled by
- TeX renderer → Diagram output: renders