[PIPE2D-327] Write code to construct 1D spectra from a combination of arc lines Created: 04/Feb/19 Updated: 06/Nov/19 |
|
| Status: | Open |
| Project: | DRP 2-D Pipeline |
| Component/s: | None |
| Affects Version/s: | None |
| Fix Version/s: | 6.0 |
| Type: | Task | Priority: | Normal |
| Reporter: | hassan | Assignee: | sogo.mineo |
| Resolution: | Unresolved | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||||||
| Story Points: | 4 | ||||||||
| Epic Link: | 1D sky subtraction | ||||||||
| Description |
|
Write code to generate 1D spectra from a combination of arc lines. The code should allow the relative intensity of the arc lines to be configurable, thus allowing fitting of element amplitudes. This would be useful for the development of a sky fitting algorithm where the relative amplitudes of OH vibrational systems may vary. |
| Comments |
| Comment by sogo.mineo [ 01/Nov/19 ] |
|
Wrote a code. It takes 10-12 hours to construct an LSF model from 4 arcs. |
| Comment by Masayuki Tanaka [ 05/Nov/19 ] |
|
Could you elaborate here? Exactly what arcs did you combine? Which lines did you use; you used the strong ones only or everything with a good flag in the line lists? As we discussed before, I expect that the computing time will increase linearly with the number of lines to fit, but 10-12 hours is clearly non-linear. |
| Comment by sogo.mineo [ 05/Nov/19 ] |
arcs = [
{"visit": 17190, "arm": "r", "spectrograph": 1, "type": ["HgI", "HgII", "ArI"]},
{"visit": 17244, "arm": "r", "spectrograph": 1, "type": ["NeI"]},
{"visit": 17424, "arm": "r", "spectrograph": 1, "type": ["KrI"]},
{"visit": 17593, "arm": "r", "spectrograph": 1, "type": ["XeI"]},
]
I use these arcs and everything in the line list with the "good" flag set. As for the long running time, I have no idea. The "10-12 hours" is shortened to 5 hours if I use 5-gaussian instead of 20-gaussian. The number of gaussians is 1/4, but the running time is not. The difference is due to the many peaks used. |
| Comment by sogo.mineo [ 06/Nov/19 ] |
|
I clipped those peaks whose heights were less than 2σ before calling scipy's minimizer. With 5-gaussian the running time reached 1 hour. |