Closed vincentsarago closed 4 months ago
Thank you for offering a permanent solution to the issue. It was indeed a challenge to identify the cause of the visualization problem. I wonder why we didn't think of integrating these additional lines of code into the existing pull request rather than starting a new one.
@amarouane-ABDELHAK
I wonder why we didn't think of integrating these additional lines of code into the existing https://github.com/cogeotiff/rio-tiler/pull/700 rather than starting a new one.
It was the fastest way for me to work on this issue and also make sure to get the results from the benchmark. If you care about ownership for this 2 lines change I could refactor this PR but I can't promise to get everything on time
@DanSchoppe this PR will change a bit the behaviour we introduced in #648
Resampling will have effect only when not overzooming
a dataset, when over-zooming it's kinda default to nearest.
This may impact you. I'm happy to discuss more to find a solution for your specific use case later
This PR re-add
width
andheight
parameters to theget_vrt_transform
(removed in #648) but will only be used when over-zooming a datasetcc @amarouane-ABDELHAK @j08lue