Fix wave elevation function.

[cmichelenstrofer]

Fixes #140. Fixes #142.

[cmichelenstrofer]

@ssolson and @rpauly18 could you review this?

• Bug fix:
  • not truncating the spectrum anymore (was being truncated based on time value, but the pd.Series was indexed by frequency).
• Speedup
  • moved calculation of B outside loop
  • replaced list comprehension + itertools with np.outer function
  • avoided creating a complex array to then take the real part only, instead create the real part directly.
  • Functionality:
  • changed default seed to None. The default behavior should be creating a random phase. A user can specify a seed for reproducibility. Should not be the other way around (where we were hardcoding a phase as the default behavior).
  • Tests:
  • There were two tests that did nothing. They would always pass. Their intention were to check that the default behavior and specifying the seed in the function signature gave the same results. But they always will even if the function is broken. I commented these out but think they should be deleted.
  • There were other tests that relied on the seed being fixed. Added a seed to those tests.
  • Examples:
  • I couldn't find an existing notebook that uses the surface_elevation function. If there isn't one, maybe we should add it to an existing example.
  • Reference
  • We need to find a good reference for this function (converting frequency domain to time domain)

[rpauly18]

@cmichelenstrofer happy to review this. Are you still working on the fix for why the tests are failing, or would you like for me to look into this?

[cmichelenstrofer]

@rpauly18 I just fixed the tests. You can go ahead and review it. Thanks!

[cmichelenstrofer]

@rpauly18 just changed it from 'draft' to 'ready to review'. I deleted the two obsolete tests that I had commented out.

[cmichelenstrofer]

Something wrong with the tests. They were passing before and nothing changed other than deleting comments. Seems the issue is not with this PR but with accessing some data at NREL needed for the tests.

requests.exceptions.RetryError: HTTPSConnectionPool(host='developer.nrel.gov', port=443): 
Max retries exceeded with url: /api/hsds/datasets/d-9db3e298-adf30787-6aa4-5dda8d-a672ee?domain=%2Fnrel%2FUS_wave%2Fvirtual_buoy%2FWest_Coast%2FWest_Coast_virtual_buoy_1995.h5&api_key=3K3JQbjZmWctY0xmIfSYvYgtIcM3CN0cb1Y2w9bf 
(Caused by ResponseError('too many 503 error responses'))

[rpauly18]

@cmichelenstrofer random failures happen sometimes because of the WPTO hindcast server being overloaded with request. I restarted the tests. If they fail again, I will investigate further.

[cmichelenstrofer]

@rpauly18 can this be merged?