uscuni / simplification

Simplification of street network geometry
Other
6 stars 0 forks source link

add outputs of manual simplification #71

Closed martinfleis closed 1 week ago

martinfleis commented 1 week ago

ref #70

@jGaboardi this may not be optimal format or placement. Feel free to move them around if you prefer different setup, I am just dumping here what I got.

jGaboardi commented 1 week ago

@anastassiavybornova Do you have any opinions on where this data should be stored? Do you think the current pathing Martin implemented here is sufficient?

anastassiavybornova commented 1 week ago

how about having subfolders for each method's output? so {fuaID}/{method}/output.gpkg, where one of the {method} subfolders will be manual (and others: parenx, osmnx, ...)

jGaboardi commented 1 week ago

how about having subfolders for each method's output? so {fuaID}/{method}/output.gpkg, where one of the {method} subfolders will be manual (and others: parenx, osmnx, ...)

Good idea. It fits the current schema. I'll pull down and make changes locally.

jGaboardi commented 1 week ago

@martinfleis @anastassiavybornova

For some FUA's there is already a manual/ directory with a .parquet inside. For example, data/1133/manual/...

martinfleis commented 1 week ago

For some FUA's there is already a manual/ directory with a .parquet inside

we should probably follow that

jGaboardi commented 1 week ago

For some FUA's there is already a manual/ directory with a .parquet inside

we should probably follow that

I will:

Any thing else to think about here?

martinfleis commented 1 week ago

That's probably it

jGaboardi commented 1 week ago

check to ensure the .parquet versions are the same data as the .gpkg ones

If I find any differences I'll report back here.

anastassiavybornova commented 1 week ago

archeology details: that was me who uploaded the .parquet ones, at some earlier iteration (for the feature matching). i got them from the gdrive and extracted only the "final version" layer before saving to parquet. but these might be superseded then by what martin uploaded?

jGaboardi commented 1 week ago

archeology details: that was me who uploaded the .parquet ones, at some earlier iteration (for the feature matching). i got them from the gdrive and extracted only the "final version" layer before saving to parquet. but these might be superseded then by what martin uploaded?

So how about I just take those ones in this PR > change directory > convert to .parquet

martinfleis commented 1 week ago

Those uploaded here are most likely exactly the same as those @anastassiavybornova uploaded before.

jGaboardi commented 1 week ago

@anastassiavybornova @martinfleis Let's get yalls eyes on this for approval before merging.