LaTeX Drawing Collection

A skill providing ready-to-use LaTeX drawing examples and guidance for creating publication-quality scientific figures using TikZ, PGFPlots, and related packages. Based on awesome-latex-drawing (2K stars), this skill covers Bayesian networks, tensor decompositions, neural architectures, time series visualizations, and more.

Overview

High-quality figures are essential for effective scientific communication. While external tools like Matplotlib or Inkscape can produce figures, native LaTeX drawings offer superior integration with the document, consistent typography, vector-quality output at any resolution, and automatic style matching with the surrounding text.

This skill equips the agent with knowledge of 30+ LaTeX drawing patterns commonly used in academic publications. Each pattern includes the required packages, a description of the drawing approach, and guidance on customization for specific research contexts.

Essential Packages

The following LaTeX packages form the foundation for scientific drawing:

TikZ (tikz)

• The core drawing package for LaTeX, providing a programming interface for vector graphics
• Supports coordinate systems, transformations, path operations, and decorations
• Required for virtually all complex scientific diagrams
• Load with: \usepackage{tikz} and relevant libraries via \usetikzlibrary{...}

PGFPlots (pgfplots)

• Built on TikZ for creating publication-quality data plots
• Supports 2D and 3D plots, error bars, fill areas, and custom markers
• Handles axis formatting, legends, and annotations
• Load with: \usepackage{pgfplots} and \pgfplotsset{compat=1.18}

TikZ Libraries

• arrows.meta - customizable arrowhead styles
• positioning - relative node placement (above=of, right=of)
• - bounding boxes around groups of nodes