Hello, it is possible to download there https://beta.source.coop/vida/google-microsoft-osm-open-buildings/geoparquet/by_country/ building footprints for entire countries for a small country like Benin 6millions shapes are available. according to https://medium.com/vida-engineering/updating-the-ultimate-cloud-native-building-footprints-dataset-6d4384cb93c4 several format are available. to my knowledge none of them can be read by M9 (pmtiles, .parquet, .flatgeobuf). some month ago I tried to use a python package though M9 (I don't remember which one) to be able to use those formats in M9, very complicated, not fully working, not fully supported. My workflow is now the following : 1. download .parquet 2. open in QGIS 3. use qGIS to export to .Sqlite 4. import the .Sqlite in M9 first question => is .sqlite the best format in term of performance to transform from .parquet to something M9 can read? then I need to select a subset of these building (like all buildings in a 10km radius around 10 cities): I did the following, but it seemed a bit non-optimal - I draw my buffers in M9,
- I did a spatial join to copy the mfd_id column of the buffers to the building drawing
- select entries where mfd_id_buffer was not null (I want to keep the whole dataset intact for later use!)
- copy selected items
- past as new table (~100 000 items)
- export to .shp for later import in Manifold8
my second question is to know if there is a better way to extract these 100k items to open in M8 ? Thank you for your help!
|