Robert Claus

- Functions who include parameters in their docstring that are not explicitly written out in the function signature encounter a **"Parameter '...' does not appear in the function signature" warning**....

- In `ml-tsne-umap-projections.md`, `from umap import UMAP` throws a `Module not Found` error. - Fixed if you do `uv pip install umap-learn`, but will throw a warning: `UserWarning: n_jobs value...

- Several files encounter errors on the line `px.set_mapbox_access_token(open(".mapbox_token").read())` because the file `.mapbox_token` cannot be found: - `hexbin-mapbox.md`, `tile-map-layers.md`, `tile-county-chloropeth.md`, `tile-scatter-maps.md`

- In`scatter-plots-on-maps.md`, the 'naturalearth_cities' dataset has been deprecated and now can be found at `https://www.naturalearthdata.com/downloads/110m-cultural-vectors/` - https://github.com/geopandas/geopandas/issues/2751

Currently each page loads Plotly.js for each example, instead of once for the page

Each page on our existing docs has a meta description and many other meta tags that need to be carried over to the new site

Right now we build the examples from jupyter notebooks here https://github.com/plotly/plotly.py/blob/main/.circleci/config.yml#L356-L388 If any notebook outputs an error the build fails and the docs won't publish We need to implement something...

Looking at https://github.com/daexs/plotly.py/pull/80/files, some of the descriptions are very short and not informative.

Probably because there are so many pages. The first time you search it can take a few seconds for any results to appear, compared with our existing search, which is...

- After converting the docstrings of each function into numpy style, the return types are not as clear and might want to be edited for clarity. For example, the following...