pydeps

Visualizes dependencies between Python modules.

Usage

Just run the pydeps.py command with a path to a directory containing Python source files.

python pydeps.py ~/Projects/CPM/src output.dot

The output is a .dot file, which can be rendered with GraphViz.

And marvel as your program's modules and their dependencies become visible:

Output Notes

• Nodes are Python modules. Edges are imports from one module to another.
• Edge thickness is proportional to the number of imports.
  • Therefore the thickness approximates the amount of coupling between adjacent modules.
• Dashed edges (not depicted) denote an import that is outside the top-level.
  • Such imports inside functions are typically used when creating a circular dependency.
  • Conditional imports at the top level are also displayed in this manner.
• Yellow modules contain a main function (via the idiom if __name__ == '__main__':).
  • Since modules with main functions are typically standalone programs or commands, it's a good idea to avoid depending on them.
• System modules are excluded from the graph.
  • A system module is defined as any module which resides outside the specified source directory.

Advanced Usage

A small number of command-line options are also supported:

Option	Description	Example
-i, --ignore-modules	Comma-separated list of modules to omit from the output graph.	crystal.xthreading, crystal.xfutures

Fun Ideas

• Create an animated history of your program's growth over time:
  • Write a program that checks out each revision of your program, runs pydeps.py, and uses the dot tool to generate an image.

Requirements

• Python 2.7
  • Earlier versions probably work too, but I haven't tested them.
• GraphViz — to render the output .dot files