You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We need to convert a geographic data cube to a cell data cube.
We need to read chunks, that are sorted by lat/lon, because the cube might be bigger than memory.
However, we want to write the cell cube sorted by cell_id, or don't we?
Does cell id order matter, there just different files on the same hard drive...
However, if we load chunks of lat/lon tiles, we must load at least everything for one cell but mybe we load even more (rectangular convex hull).
Maybe it is better to do this:
for cell_id in 1::num_cells:
boundariy = get_boundary(grid, cell_id)
chunck = get_chunk(boundary, geo_cube)
cell_cube[cell_id] = mean(chunck)
Currently, we load rectangular lon/lat tile chunks and use matrix vector multiplication.
The text was updated successfully, but these errors were encountered:
We need to convert a geographic data cube to a cell data cube.
We need to read chunks, that are sorted by lat/lon, because the cube might be bigger than memory.
However, we want to write the cell cube sorted by cell_id, or don't we?
Does cell id order matter, there just different files on the same hard drive...
However, if we load chunks of lat/lon tiles, we must load at least everything for one cell but mybe we load even more (rectangular convex hull).
Maybe it is better to do this:
Currently, we load rectangular lon/lat tile chunks and use matrix vector multiplication.
The text was updated successfully, but these errors were encountered: