Kyle Barron

I don't have plans to work on this repo myself, but PRs are welcome. It probably wouldn't be too hard to update this PR with changes in `master`

```py import geopandas as gpd from keplergl_cli import Visualize gdf = gpd.read_file('/Users/kyle/Downloads/nationalpark_orig.geojson') vis = Visualize() vis.add_data(gdf) html_path = vis.render(open_browser=False) with open(html_path) as f: source = f.read() from IPython.display import IFrame...

> but this does not work as intended The lz4js API works on arrays. You can't pass a string to it. This works: ```js > lz4js.decompress(lz4js.compress(new Uint8Array([1, 2, 3, 4])))...

Did you also change the path within `stata-png-fixed.sh`? https://github.com/kylebarron/stata-png-fix/blob/a61139113eb3fd29bdaae9f8fdcb71c0117e810e/stata-png-fixed.sh#L3 I didn't spend much time on this so things are hardcoded to my paths.

> Though removing `mode="json"` might have regressions for other data types, like dates? I'm assuming this will end up being a problem... It looks like I got this to work...

It seems like you either need to manually handle `bytes` if setting `model_dump(mode="json")` or you'd need to manually handle dates if setting `model_dump()`, so not sure which is easiest 🤷‍♂️

To play devil's advocate, `datetime.datetime` is a python builtin too? It's hard because it seems the Jupyter comm requires _almost JSON_, in that everything but `bytes` needs to be json-serializable?

It doesn't look like there's a way to do this in the CloudFormation template, so maybe just add a note to the README

This project was only for US-based contours, and even then only generated contours for western US. The only global contours mbtiles that I know of that you can download is...

By the way, it looks like it makes 512x512 tiles by default.