Feat/convert viirs to module

[butchtm]

• adds a povertymapping.nightlights module
• adds a povertymapping.nightlights.generate_nightlights_feature method that adds avg_rad_mean,avg_rad_max , avg_rad_min,avg_rad_median columns to the aoi.
• downloads the nightlights tif file from the EOG site, unzips and and it in the cache directory ~/.geowrangler/nightlights/global
• creates a clipped raster file using the total bounds of the aoi and caches it in the cache directory ~/.geowrangler/nightlights/clip
• adds a method get_eog_access_token to create and store the eog access token using the user's registered username and password (instructions and links to registration page are included in the test notebook).
• implements #66

[review-notebook-app[bot]]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[butchtm]

Unlike the ookla and osm datasets, nightlights uses raster files so I don't actually load the raster file, - I just create a clipped raster file from the global file using the total bounds of the aoi -- this is actually pretty fast (about 15msecs). The slowest parts of the process are:

• download the global nightlights data (only done once for each year) - 30-45 mins - 11GB download
• unzip downloaded file (11-20 secs) - done once per download
• raster zonal stats computation (depends on aoi) (around 20 secs)