Matplotlib Backend for DXF

Overview

This is a Matplotlib backend that enables Matplotlib to save figures as DXF drawings. DXF is a drawing format commonly used by Computer-Aided Design (CAD) tools.

This package builds on the ezdxf package by Manfred Moitzi: ezdxf on Bitbucket

Installation of dependencies

The package dependencies can be installed with uv

uv sync --upgrade
source .venv/bin/activate

Usage

To use this backend, you first need to register it with Matplotlib:

import matplotlib
from mpldxf import backend_dxf

# Matplotlib registration takes a FigureCanvas *class* (no kwargs), so use the
# factory to configure options like subplot sub-blocks.
FigureCanvas = backend_dxf.make_figure_canvas(use_fm_layers=False, use_subplot_blocks=True)
matplotlib.backend_bases.register_backend("dxf", FigureCanvas)

Then, you can save a figure as a DXF file:

from matplotlib import pyplot as plt
plt.plot(range(10))
plt.savefig('myplot.dxf')

Feature flags

mpldxf can optionally write each subplot into its own DXF block (nested under a single main_plot block) to keep modelspace clean and make downstream CAD reuse easier.

• Default: disabled
• Enable (legacy “write directly to modelspace” behavior): pass use_subplot_blocks=True via backend_dxf.make_figure_canvas(...).

Warning

This package is a work in progress. Not all Matplotlib plot types will render correctly, and text alignment and sizing in particular may require adjustments.