I am trying to fill NODATA although I do not know what values they are represented in.
So running:
gdal_fillnodata.py on a raster file, did not give the desirable results, as not all NODATA cells are filled.
The command I picked from QGis (that is directly using gdal gdal_fillnodata.py) then tried to run in Windows bash like:
FOR %i IN (C:\Users\UBU\Dowloads\NODATA\*.tif) DO "C:/Program Files/QGIS 3.12/bin/python3.exe" "C:/Program Files/QGIS 3.12/apps/Python37/Scripts/gdal_fillnodata.py" -md 10 -b 4 -of GTiff "%i" "%i.tiff"
But this does not change anything I guess.
Result:
Using Raster object from acrpy provided by Esri, I am testing nan values in the file in input and in output:
I am trying to fill NODATA although I do not know what values they are represented in. So running:
gdal_fillnodata.py
on a raster file, did not give the desirable results, as not all NODATA cells are filled. The command I picked fromQGis
(that is directly using gdalgdal_fillnodata.py
) then tried to run in Windows bash like:FOR %i IN (C:\Users\UBU\Dowloads\NODATA\*.tif) DO "C:/Program Files/QGIS 3.12/bin/python3.exe" "C:/Program Files/QGIS 3.12/apps/Python37/Scripts/gdal_fillnodata.py" -md 10 -b 4 -of GTiff "%i" "%i.tiff"
But this does not change anything I guess.
Result:
Using Raster object from acrpy provided by Esri, I am testing nan values in the file in input and in output:
np.count_nonzero(np.isnan(RasterToNumPyArray(Raster('s5p_no2_20180917T164548_20180923T182508.tif'))))
600np.count_nonzero(np.isnan(RasterToNumPyArray(Raster('20180917T164548_20180923T182508.tif.tiff'))))
50So I still have some NODATA cells. Is this normal behavior ? and why if yes.