Repository navigation
First public release.
export-igorgraph turns a matplotlib figure into an editable Igor Pro graph:
export_igor(fig, outdir, name) writes one HDF5 file (data + Igor drawing commands),
and LoadPythonFigure in Igor draws it.
See README (EN) / README (JP) for details.
Claude skill
export-igorgraph-public.skill (attached in this page) is a skill file for Claude.ai (upload it in the skill settings) and Claude Code (unzip it into ~/.claude/skills/). It bundles the Python package and keeps your default formatting in export_igorgraph_style.json.
Highlights
- Lines (solid and dashed), 20 marker shapes (5 of them approximated, with a warning), error bars with caps, log axes (10^n labels), grid, tick direction, legends, labels
- Scatter plots with one color and size (per-point colors/sizes: the first value is used, with a warning)
- Images, color maps, framed color scale
- Contours drawn as real Igor contours (
AppendMatrixContour/ModifyContour) from the original Z matrix (matplotlib 3.8+) - Main formatting via
style=or a style file:tick,mirror,standoff,grid,width,height,font,gFont - The Igor loader runs only commands that match a strict allow-list and rejects the whole file otherwise
- Anything Igor cannot reproduce is reported in
report.warnings, never dropped silently
Install
pip install export-igorgraph(orpip install git+https://github.com/nlyamada/export-igorgraph)- Open
ExportIgorGraphLoader.ipf(attached in this page), or copy to Igor procedure folderDocuments\WaveMetrics\Igor Pro * User Files\Igor Procedures\and restart Igor. - In Igor:
Macros > LoadPythonFigure
Status
- Verified on Igor Pro 8.04 / Windows only. Other Igor versions and macOS are unverified — reports welcome.
- Python 3.9+. CI runs matplotlib 3.6.3, 3.8.4, 3.9.4 and 3.10.7 (contours need 3.8+).
See the CHANGELOG and the documentation.