Re-project match just uses the same CRS, transform, and shape of the other raster:
\n
\n rioxarray/rioxarray/raster_array.py\n
\n\n Line 552\n in\n 6b768ec\n
\n| \n | reprojected_data_array = self.reproject( | \n
Examples referenced in solution:
\n\nSolution:
\nimport rasterio\nimport rasterio.vrt\nimport rioxarray\n\nwith rasterio.open(\"large.tif\") as src, rasterio.open(\"template_large.tif\") as template:\n with rasterio.vrt.WarpedVRT(\n src,\n crs=template.crs,\n transform=template.transform,\n height=template.height,\n width=template.width,\n ) as vrt:\n rds = rioxarray.open_rasterio(vrt, lock=False, parse_coordinates=False)\n rds.rio.to_raster(\"output.tif\", tiled=True, windowed=True)
-
|
I am trying to reproject a large raster, using a second raster as a template. This works fine on smaller rasters, but crashes with larger ones. I think I probably need to use window'd read and writes but I'm unsure how to implement this or if it's the correct solution. Thanks. |
Beta Was this translation helpful? Give feedback.
Replies: 3 comments 6 replies
-
|
Re-project match just uses the same CRS, transform, and shape of the other raster: rioxarray/rioxarray/raster_array.py Line 552 in 6b768ec Examples referenced in solution: Solution: import rasterio
import rasterio.vrt
import rioxarray
with rasterio.open("large.tif") as src, rasterio.open("template_large.tif") as template:
with rasterio.vrt.WarpedVRT(
src,
crs=template.crs,
transform=template.transform,
height=template.height,
width=template.width,
) as vrt:
rds = rioxarray.open_rasterio(vrt, lock=False, parse_coordinates=False)
rds.rio.to_raster("output.tif", tiled=True, windowed=True) |
Beta Was this translation helpful? Give feedback.
-
|
Thanks for the help @snowman2 . Not very experienced with this library so struggling to implement your suggestion a bit. Looking at the URL you gave |
Beta Was this translation helpful? Give feedback.
-
|
i.e. I can't see from the example how I can use a 'template' raster, and/or how to write out the result. |
Beta Was this translation helpful? Give feedback.
-
|
I used an AI bot to come up with this. Going to test it now. |
Beta Was this translation helpful? Give feedback.
-
|
@JimShady I added a code example to reference. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks so much @snowman2 . Will give it a whirl over the weekend. |
Beta Was this translation helpful? Give feedback.
-
|
reproject match is like this template_raster = rioxarray.open_rasterio('template_raster.tif')
useraster = rioxarray.open_rasterio('useraster.tif')
finalraster = useraster.rio.reproject_match(template_raster)
finalraster = finalraster.assin_coords({"x":template_raster.x, "y":template_raster.y}) #so they match exactly |
Beta Was this translation helpful? Give feedback.
-
|
Thanks @RichardScottOZ . Unfortunately I've still not managed to figure out how to do this with large rasters. I think I'm going to resort to 'raw' gdal on the command line via os.system() commands. |
Beta Was this translation helpful? Give feedback.
-
|
This is GPT's answer, I would like to know why the code execution to reprojected_ds only takes 17.3s, however using rio.to_raster() over 40 minutes still for completion, is it because my parameter setting is not right? |
Beta Was this translation helpful? Give feedback.
Re-project match just uses the same CRS, transform, and shape of the other raster:
rioxarray/rioxarray/raster_array.py
Line 552 in 6b768ec
Examples referenced in solution:
Solution: