Codecov Report

Merging #285 (b44f998) into master (a1529cc) will not change coverage.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master     #285   +/-   ##
=======================================
  Coverage   90.64%   90.64%           
=======================================
  Files          37       37           
  Lines        5943     5943           
=======================================
  Hits         5387     5387           
  Misses        556      556           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a1529cc...b44f998. Read the comment docs.

@eecarres this looks like a good way to go, although I recommend to introduce a more generic approach, something like, rasterio_kwargs which accepts a dictionary of arguments passed to rasterio.warp.reproject

@arielze Whatever you find the most convenient. What I saw in the telluric code was that direct calls to rasterio reproject were done with the option of giving all the parameters explicitly, that's why I tried to "continue with the trend". As you can see here the rasterio call also has a kwargs, but the rest of the parameters (except from the two I added) are explicitly defined in the telluric calls.

As I said, you decide =D

I do agree with @arielze. These two parameters are too low level and we don't deal with them in the source code. So I think it worth to create a new **kwargs argument for merge_all. In this case we don't need to copypaste docstring with description of these parameters into myriads of internal method and also it allows us to use new arguments that might be added in the future to rasterio.