Hi, thanks for your response. Let me provide some background. I have two sets of images that capture the same scene but were taken using separate cameras. This causes some degree of GPS discrepancy

Hi Sean I can't find the dev-group to discuss this issue, where can I find it? Thanks, Guy

Is this the right place to report a bug? I can open an issue on github if you like. I'm reading a NITF, and getting this error. Should add 'JPEG2000' to the Compression enum? In[14]:

Hi, This is indeed a bug. I would be much obliged if you reported it in the Rasterio issue tracker. I haven't used NITF in a long time. I see in https://www.gdal.org/frmt_nitf.html that there's a

Hi, What an interesting problem. The reproject() function is mainly for cartographic reprojection, but it might be useful in your case. The row, col, x, and y for a GroundControlPoint need to be

I am trying to recreate gdal_calc for simple calculation using below code. import numpy import rasterio def merge_bands(calculationType, band1Image, band2Image, outputFilePath): # Read raster bands

Hi, Can you also include the expressions that you're using with gdal_calc? Without them we can only guess at the cause of the difference in the results. I don't see any big problems with the way

Hi Sean, Sure, I will try those changes you have mentioned. here is the gdal_calc expression I used : gdal_calc.py -A T1_B5.TIF --A_band=1 -B T1_B4.TIF --B_band=1 --outfile=ndvi_gal_1.TIF

Hi this is a little bit of a strange question about how rasterio is using the GDALRasterizeGeometries function. This c-function is called in _features.pyx here: with InMemoryRaster(image=image,

You're right, the transform that Rasterio passes to GDALRasterizeGeometries is not used. Worse, there looks like risk of corruption. I'm going to change Rasterio to pass NULL instead. When

Yeah I thought so. FYI the code you linked is the same one I embedded above. It looks like it really is just relying upon the in memory raster already having a configured transform associated with it.

I've been looking for the best way of processing a large raster files while achieving both maximum processing parallelism/speed and a low memory footprint. To achieve a low memory footprint, is enough

Pablo, I think your approach would make a great example and addition to the concurrency docs. Being able to do input, output and computation in parallel is a huge benefit. The locks seems effective at

Hi Pablo, I really appreciate your analysis of the concurrency example! Things have changed quite a bit in the newest versions of Python since it was written and it needs an update. While researching

Hi I forked and am trying to add support for rasterio and rio to handle files on Google Cloud Storage. I made changes to path.py and added remote schemes to include gs:// type urls similar to AWS S3

Hi, Your GDAL library needs to be built on a curl library that is itself built with SSL support. I think that the GDAL and curl packages in most distros will meet this requirement, but if you are

Hi Sean! Oh, that's really interesting. Thanks for sharing this. I'm going to do my own tests and then update the PR I made with the example.

Actually, this warning is GDAL specific, and has nothing to do with curl (but of course curl without SSL support is rather useless). The service account authentication method of Google OAuth2

I didn't build gdal but I am using the base gdal image from quay.io/mojodna/gdal:v2.3.2. s3 access is working perfectly fine. If GDAL was not built SSL, AWS S3 wouldn't have worked correct? Also is

Thanks for the clarification, Even!-- Sean Gillies