Split file by day?

[Mike-Honey]

Would this project work better if we split the input shapefile into individual days?

I had a thought - while the dataset for a month or several months is quite large, the data for a single day would be a lot smaller and easier to process (simplification, tippecanoe etc). This does assume that the data for previous days doesn't change.

For the mapbox visual to work, the script would need to append all the days in the desired range (e.g. last 3 months) into one mbtiles file before uploading to mapbox. I'm not sure if you can append mbtiles files, that might need to happen upstream.

[Mike-Honey]

I found this in the tippecanoe doco:
https://github.com/mapbox/tippecanoe#input-files-and-layer-names

It looks like the -l option would combine multiple geojson files into one layer in a single mbtiles output file. So the python script would need to build a long command line of all the daily files.

[Mike-Honey]

The output file needs a unique attribute - a ROW_NUMBER that restarts each day wouldn't work when the Mapbox visual tries to filter the data. Could you create a concatenation of the Date & ROW_NUMBER?