How does the skill handle diagram layout?

It primarily uses automated layout engines like Graphviz's 'dot' or 'neato' to calculate optimal node positioning, minimizing edge crossings and ensuring clear readability.

What types of neural networks are supported?

It supports a wide range including Transformers (encoder-decoder), CNNs, RNNs, LSTMs, and custom feed-forward architectures with automated layer stacking.

Which Python libraries does this skill use?

It utilizes Graphviz (via pygraphviz or pydot), Matplotlib, Schemdraw, NetworkX, and the Diagrams library to ensure the best output for different diagram types.

Can I use these diagrams directly in LaTeX?

Yes, the skill generates standalone PDF, SVG, and EPS files and provides the specific LaTeX code snippets to include them in your document via the graphicx package.

Scientific Schematics & Diagrams

Name: Scientific Schematics & Diagrams
Author: Zehong-Wang

byZehong-Wang

0•

データサイエンスとML

Generates publication-quality scientific diagrams, neural network architectures, and technical flowcharts using specialized Python libraries.

The Scientific Schematics skill empowers researchers and engineers to programmatically create professional vector graphics for academic papers and technical documentation. By leveraging industry-standard Python libraries like Graphviz, Matplotlib, and Schemdraw, it automates the generation of complex neural network architectures, circuit diagrams, biological pathways, and system flowcharts. This skill follows a specialized workflow that ensures all diagrams are stored as high-resolution SVG or EPS files in dedicated project folders, facilitating seamless integration into LaTeX documents or technical posters without the need for manual drawing tools.

主な機能

01Automated generation of Transformer, CNN, and RNN architecture diagrams

02Automated quality verification and organized figure folder management

03Engineering-standard circuit diagram generation using Schemdraw

04Professional flowchart creation for methodology and process workflows (CONSORT/PRISMA)

05Export to high-resolution vector formats (SVG, EPS, PDF) for publication

060 GitHub stars

ユースケース

01Designing complex system data flows and software architecture diagrams

02Generating standardized clinical trial flowcharts and experimental methodologies

03Visualizing deep learning architectures for research papers and presentations

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add zehong-wang/kosmos scientific-schematics

For use in Claude.ai and ChatGPT

Download Skill