search

NREL / reV

Renewable Energy Potential (reV) Model
https://nrel.github.io/reV/
BSD 3-Clause "New" or "Revised" License
107 stars 24 forks source link

refactor tech_mapping #281

Closed MRossol closed 3 years ago

MRossol commented 3 years ago

@grantbuster I had to update the baseline techmap due to removing the ability to set the distance threshold. Now it's computed with a user-defined dist_margin (default 1.05). Plotted below is the difference between the old baseline and the new techmap (black dots). The new techmap is slightly more conservative (i.e., more -1 pixels)

Figure_1

grantbuster commented 3 years ago

The map looks great! a bit of difference on the edges of the resource domain is fine. I think we should plot up the full techmap at some point to make sure there's no gaps though. The test case is pretty ideal without terrible skew to the resource grid.

MRossol commented 3 years ago

So the -1 issue was due to Travis' lat, lons.

Performance on the test domain: Master: In [4]: %%time ...: ind = TechMapping.run(EXCL, RES, dset=None, max_workers=2) ...: ...: CPU times: user 166 ms, sys: 188 ms, total: 354 ms Wall time: 4.89 s

New branch: In [2]: %%time ...: ind = TechMapping.run(EXCL, RES, dset=None, max_workers=2) ...: ...: CPU times: user 78.4 ms, sys: 68.4 ms, total: 147 ms Wall time: 3.89 s

Not a great test case, but still faster!

grantbuster commented 3 years ago

Sweet! if it works, its faster, and its more simple then hell yes.