Kaleido is a cross-platform library for generating static images for Plotly's visualization library.
After installing it, you can use fig.write_image("filename.png") to save a plot to a file.
The original version of kaleido included a custom build of the Chrome web browser, which made it very large (hundreds of megabytes) and proved very difficult to maintain. In contrast, this version depends on choreographer, a lightweight library that enables remote control of browsers from Python. When you ask kaleido to create an image, it uses choreographer to run a headless instance of Chrome to render and save your figure. Please see choreographer's documentation for details.
The new version of kaleido is a work on progress; we would be grateful for help testing it and improving it. If you find a bug, please report it in our GitHub repository, and please include a minimal reproducible example if you can.
It would also be very helpful to run the script
src/py/tests/manual.pyand attach its zipped output to your bug report. This will give us detailed information about the precise versions of software you are using and the platform you are running on, which will help us track down problems more quickly.
You can install kaleido from PyPI using pip:
$ pip install kaleido
Versions 4.9 and above of the Plotly Python library will automatically use kaleido for static image export when kaleido is installed. For example:
import plotly.express as px
fig = px.scatter(px.data.iris(), x="sepal_length", y="sepal_width", color="species")
fig.write_image("figure.png", engine="kaleido")See the Plotly static image export documentation for more information.