vtsuperdarn / davitpy

DEPRECATED The DaViT Python project
http://vtsuperdarn.github.com/davitpy/
GNU General Public License v3.0
37 stars 59 forks source link

elevation angle recalculation bug #359

Closed aburrell closed 4 years ago

aburrell commented 6 years ago

Fixed a bug in the elevation angle recalculation that didn’t check to see if the interferometer cables had been switched or not.

Found after consultation with @egthomas, who was working on an IDL implementation of Simon's elevation angle calculation.

aburrell commented 6 years ago

Test!

import davitpy; import datetime as dt stime = dt.datetime(2014, 1, 1); etime = dt.datetime(2014, 1, 1, 1) rad = 'han'; hard = davitpy.pydarn.radar.site(code=rad, dt=stime) file_type='fitacf' rad_ptr = davitpy.pydarn.sdio.radDataRead.radDataOpen(stime, eTime=etime, radcode=rad, fileType=file_type) bm = rad_ptr.readRec() out = davitpy.pydarn.proc.fov.calc_elevation.calc_elv(bm, hard=hard) hard.phidiff = -1 out_neg = davitpy.pydarn.proc.fov.calc_elevation.calc_elv(bm) print out[0] [ nan 17.62225948 14.08461472 14.13857093 nan 22.79720948 22.50723233 nan nan nan nan 28.97115884 25.81253934 nan 1.3868721 30.12459279 1.76833418 19.39947718 6.18414091 25.60871328 nan nan nan nan nan] print out_neg[0] [ nan 27.70894511 29.71038192 29.68393623 nan 23.59279877 23.87010294 nan nan nan nan 15.50954806 20.26437913 nan 11.48070041 13.2058227 11.41097333 26.47591919 9.58083037 20.51899942 nan nan nan nan nan]

egthomas commented 6 years ago

Simon and I were talking about this yesterday and he doesn't actually have a definitive answer on how the phidiff parameter should affect the elevation angle determination. It looks like that parameter was only ever set for a brief period at the Goose Bay radar (when there may not even be any available xcf data); the only other radar with a non-zero phidiff is South Pole but the PI has indicated that is a mistake in the hdw file. Sorry for causing the extra work on your end!

aburrell commented 6 years ago

Well, it's good to have it just in case! I couldn't find any data with the flag set to -1, which was comforting (so nothing was processed wrong).

ksterne commented 4 years ago

This pull request never got going here even though the work is certainly good to consider. Closing this now as this repo is being deprecated.