Below is a summary of the changes I made! At the end, I describe 2 issues that still exist but do not prevent successful running of the code.

ee_extraction

made an additional subfolder called ee_extractions so that multiple processes can use the same directory but subfolders like exports/foldrers/etc are saved in unique trees

getGeometry

No longer necessary; usurped by getLocation
Moved to scratch → old_code

getLocation

FOR POINTS:

for points exported as ee, return a list of features instead of a collection (for easier iteration in extracting them one by one). The conversion to feature collection line is commented out.
changed bounding box to export as ee.Geometry instead of ee.Feature(ee.Geometry) because that is what .filterbounds() wants

FOR WATERSHED:

in the ee conversion section, I added back in an elif for watersheds and left the multipolygon 'else' category untouched. I'm not totally sure how much of the watershed part is necessary/redundant, or if it really needs to be a feature collection instead of a feature.

IN GENERAL:

added output_directory and sub_directory so that files go to the right places

Continuing Issues:

[ ] ISSUE 1:

In getLocation, line 68 userWarning: Geometry is in a geographic CRS. Results from 'buffer' are likely incorrect. Use 'GeoSeries.to_crs()' to reproject geometries to a projected CRS before this operation.... full note:

*USGS Basin imported at Elder C Nr Branscomb CaCRS: epsg:4326
/Users/ericamccormick/Documents/GITHUB/rock-moisture/code/api_extractions/getLocation.py:68: UserWarning: Geometry is in a geographic CRS. Results from 'buffer' are likely incorrect. Use 'GeoSeries.to_crs()' to re-project geometries to a projected CRS before this operation.
bbox_gdf = gpd.GeoDataFrame([1], geometry=[box(*sites.buffer(0.05).total_bounds)], crs=sites.crs)*

[ ] ISSUE 2:

In ee_extractions, LINE 68, when I try to run direct_storage...

subprocess.call('cp ' + os.path.join(args.point_csv) + ' ' + os.path.join(args.output_directory, 'ee_extractions', 'settings', 'export_coords.csv'), shell=True) ← DOESNT WORK

subprocess.call('cp ' + os.path.join(args.output_directory, 'ee_extractions', 'settings', 'export_coords.csv'), shell=True) ← DOES WORK

Traceback (most recent call last):
  File "/Users/ericamccormick/Documents/GITHUB/rock-moisture/code/api_extractions/ee_extractions.py", line 75, in <module>
    main()
  File "/Users/ericamccormick/Documents/GITHUB/rock-moisture/code/api_extractions/ee_extractions.py", line 67, in main
    subprocess.call('cp ' + os.path.join(args.point_csv) + ' ' + os.path.join(args.output_directory, 'ee_extractions', 'settings', 'export_coords.csv'), shell=True)
  File "/Applications/anaconda3/envs/ee/lib/python3.9/posixpath.py", line 76, in join
    a = os.fspath(a)
TypeError: expected str, bytes or os.PathLike object, not NoneType

Ok, I pulled in your changes and made some follow up edits. Looking good! There were a few comments above, but the main issue is I can't run the direct storage plotting code - throws an error around line 20 that i commented on in line in my push back to my main branch.

For your issue 1 above - yeah, I checked it and the bounding box that gets printed on the map has always looked fine so i just ignored it, but we could do a transform to and from a projected space. I think that there's a function in geopandas to estimate the most appropriate epsg code for a lat/long, but we would have to dig that up.

For the second one, i need to check what's happening there... It was working fine for me before and the cp call should want both what to copy and where to put it, so need to check on that.

In get_data_df.py I commented out the watershed flag - mine works both ways with the same function now, let me know if that doesnt work for you though and we can figure out a resolution.

Finally, I think that for the extraction export, we should be able to reduce the amount of redundant code by doing small pieces that behave differently based on watershed or not, but that's low priority. Just might help eventually to keep the two versions from getting out of sync.

kdchadwick / ee-extractions