Calibration misc #325
Calibration misc #325
Conversation
| if total(matched_times_opt1 < 1) GT total(matched_times_opt2 < 1) then matched_times = matched_times_opt1 | ||
| if total(matched_times_opt1 < 1) LT total(matched_times_opt2 < 1) then matched_times = matched_times_opt2 |
There was a problem hiding this comment.
what if neither are very good? should we error or warn if the best one is still way off?
There was a problem hiding this comment.
I've included two errors to be explicit, one if no times matched and one if the times matched do not equal the number of timesteps in the data.
| ; In each matched timestep, match the baselines and fill a new, matched model array | ||
| for time_i=0, data_n_time-1 do begin | ||
| ;suba is the subset of indices in the first array that are also in the second. | ||
| ;subb is the subset of indices in the second array that are also in the first. | ||
| match, data_baseline_index[time_i*data_nbaselines:(time_i+1)*data_nbaselines-1], $ | ||
| model_baseline_index[matched_times[time_i]*model_nbaselines:(matched_times[time_i]+1)*model_nbaselines-1], $ | ||
| suba, subb | ||
| for pol_i=0, obs.n_pol-1 do (*matched_model[pol_i])[*,suba+time_i*(data_nbaselines)] = (*vis_model_arr[pol_i])[*,subb+matched_times[time_i]*model_nbaselines] | ||
| endfor |
There was a problem hiding this comment.
Ahh, so this takes care of the ones that don't match? what exactly happens if there are times that don't match?
There was a problem hiding this comment.
I've added an extra comment to be a little more clear.
Any unmatched transferred model times are discarded. The idea here is that you have a full model for all times, and you are matching it to imperfect, flagged data.
You may have noticed that I did not force timestep matching if the number of times is the same between the data and the model. I want to keep it somewhat flexible in case people are working with simulations. Ideally everyone fills in the metadata for simulations, but I'm not going to force them.
|
Added an extra option for the newer version of WODEN (which includes a better JDate convention after realizing there was such a huge difference). |
These are miscellaneous calibration changes:
All changes have been confirmed to work in PS.