-
Notifications
You must be signed in to change notification settings - Fork 8
Jan 18, '22
Avi Vajpeyi edited this page Jan 25, 2022
·
4 revisions
-
Raw Data handling:
- Check cadence: 2min cadence
- Use only SPOC data:
- Restrict
lk
search toauthor=SPOC
- Filter ExoFop candidates to only those with
source=SPOC
- this reduces TOIs 5K -> 1.7K (https://github.com/dfm/tess-atlas/pull/168)
- Source = Parameter Source Pipeline - This column lists the data analysis pipeline which initially identified the planet candidate
- Pipeline = Detection Pipeline(s) = This column lists the pipeline which generated the current set of parameters for a given TOI.
- https://tess.mit.edu/news/toi-catalog-now-includes-the-detection-pipeline-for-each-candidate/
- Restrict
- Adjusted
quality-bitmask
tohard
instead ofdefault
(not usinghardest
as not recommended) - Adjusted
lk.remove_outliers
to prevent sigma clipping (see https://github.com/dfm/tess-atlas/pull/171)
-
phase plot issue: GP not subtracted from flux:
- Compute
mu=GP.predict(residuals)
in post-proc and subtract out from flux (see https://github.com/dfm/tess-atlas/issues/173)
- Compute
We didn't finish going over all the TOIs in the last meeting:
- Ran analysis on all TOIs, documented here: https://github.com/dfm/tess-atlas/issues/166
- Some TOI fits look strange, documented here: https://github.com/dfm/tess-atlas/issues/167
- Results:
- What diagnostic plots can I make?
- Shall we try to see if the initial guess is good? (https://github.com/dfm/tess-atlas/issues/101)
- Manually run TOI 178 (multiplanet) TOI 237 (potentially multi-modal fit?)
- Make and save diagnostic plots (binned data, corner, traces, SNR?)
- Look at phase-plots and divide erroneous fits in categories
- eg. wrong period init https://github.com/dfm/tess-atlas/issues/101
- use the quality flag --> hardest (not a big deal if we delete extra data)
- remove EXOFOP
source=SPOC
filter once complete with debugging