154 Scientific Diagrams

Illustrating concepts in physics, chemistry and machine learning.

janosh.github.io/diagrams allows searching, sorting, opening in Overleaf and downloading figures (PDF/SVG/PNG) from this collection.

Have a TikZ/CeTZ diagram you'd like to share? Submit a PR with a .tex or .typ and a corresponding metadata .yml file in the assets/ directory and add yourself to the citation.cff file.

2d Convolution	Amplitude Modulation
Angular Momentum Quantization	Atomistic GPU Batching
Autoencoder	Ball Tree
Basis + Lattice	Bloch Sphere
Bose Einstein Distribution	Bose Einstein Distribution 3d
Branch and Bound	Branch Cuts 1
Branch Cuts 2	Burrows-Wheeler Transform
Cartesian vs Polar Coordinates	Chain Homology
Change of Variables	Closed String Topologies
Complex Sign Function	Complex Sign Plane
Concave Functions	Convex Functions
Convex Hull of Stability	Convolutional Autoencoder
Critical Temperature	CycleGAN
Cylinder to Plane	De Bruijn Graph
Deep Belief Network	Denoising Variational Autoencoder
DFT Choices	DFT Jacob's Ladder
DFT MLFF CFF Speed Accuracy Transfer	Disk to Plane
Divergence	DNA Double Helix
Dropout	Ergodic
Euler Angles	Fermi-Dirac Distribution
Ferroelectric Response	Feynman Diagram 1
Feynman Diagram 2	Feynman Diagram 3
Feynman Diagram 4	Feynman Diagram Loop
Feynman Diagram Loops	Feynman Diagram One Point
Feynman Diagram Propagator Loop	Feynman Diagrams Loop Regulator
Four Vs of Data	Frequency Modulation
Gas Pressure on Wall	Generative Adversarial Network
Geometric Bayes	GNN Node Aggregation
Grand Canonical Occupation Fluctuations	Graph Attention Layer
Graph Convolution	Graph Isomorphism
Gravitons	Gun Tackle
Hamiltonian Graph	Harmonic Oscillator Energy vs Angular Frequency
Harmonic Oscillator Energy vs inverse Temperature	Heatmap
Higgs Potential	High Entropy Alloy
In-Flight Auto-Batcher	Iron Lattice Ferrox Logo
Isotherms	Jensens Inequality
k-Nearest Neighbors	k-Space
Kohn Sham Cycle	Light
Long Short-Term Memory	M-Theory
MADE	Masked Autoregressive Flow
Materials Informatics	Materials Informatics Challenges
Matsubara Contour 1	Matsubara Contour 2
Matsubara Contour 3	Matsubara Contour 4
Matsubara Contour 5	Matsubara Contour Deformation
Maximum Flow Problem	Maxwell Boltzmann Distribution
Message Passing Neural Network	Mexican Hat
ML Activations	Momentum Shell
MOSFET	Multilayer Perceptron
NMOSFET Amplifier Circuit	Normalizing Flow
Normalizing Flow Coupling Layer	Open String Topologies
Operator Orderings	Organic Molecule
Otto Cycle	Periodic Table
Physics Mindmap	Plane to Torus
Plate Capacitor	Potential Triangle
Propagator Fluctuations	QFT Propagator Poles
QFT Propagators	Cost vs Accuracy in Quantum Mechanical Simulations
Random Forest	Regular vs Bayes NN
Reinforcement Learning Greedy Policy	RG Flow in Theory Space
Risk-Opportunity Matrix	RNVP Affine Coupling Layer
ROC Curve	Roost Update
Sabatier Principle	Saddle Point
Scattering Detailed Balance	Seebeck Effect
Self Attention	Semi-supervised Learning
Shortest Path in Weighted Graph	Sierpinski Triangle
Signal Sampling	Single-head attention
Skip Connection	Sparse Autoencoder
Spontaneous Magnetization	Statistical Energy Distributions
Sublattice Points	Supervised Learning Setup
Tanh	Thermodynamic Ensemble transforms
Thermodynamic Potentials Venn Diagram	θ-β-M Diagram
Thomson Scattering	Timed Event Graph
Tori	Torus
Torus Fundamental Domain	Train-Test Split
Two Point No Cutoff	Two-point Propagator Flow
Unregularized Propagator Diagrams	Variational Autoencoder
Web Graph	Wetterich Equation
Wyckoff Positions	zT vs n

Scripts

Files in /scripts render and convert the .typ and .tex files in /assets to various formats:

• low + high-res PNG
• PDF
• SVG

To run the scripts requires the following dependencies:

• pdf-compressor (pip install pdf-compressor)
• gs (GhostScript) (optional, worse compression but needs no API key so less setup than pdf-compressor)
• pdf2svg (brew install pdf2svg)
• convert (part of ImageMagick)
• pngquant (brew install pngquant)
• zopflipng (brew install zopfli)

To run pdf-compressor directly or to use it as part of the render-(typst|tikz).py pipeline, you need a free public API key from https://developer.ilovepdf.com. Pass it to pdf-compressor with:

pdf-compressor --set-api-key project_public_7c854a9db0...

📖   How to cite

You can cite the Zenodo record using the following BibTeX entry:

@software{riebesell_diagrams_2020,
  title = {Collection of scientific diagrams},
  author = {Riebesell, Janosh and Bringuier, Stefan},
  date = {2020-08-09},
  year = {2020},
  doi = {10.5281/zenodo.7486911},
  url = {https://github.com/janosh/diagrams},
  note = {10.5281/zenodo.7486911 - https://github.com/janosh/diagrams},
  version = {0.2.0},
  urldate = {2023-01-01}, % optional, replace with your date of access
}