Caltech-IPAC / hiresprv

Keck HIRES PRV pipeline service access
4 stars 0 forks source link

RV curve file contains all results for an object #39

Open cgelino opened 4 years ago

cgelino commented 4 years ago

I want to run the template, rv, and rvcurve for a single night, but when I look at the RV curve file for that object, it contains all of RV measurements for that object that were previously run and dumped in my workspace. There should be a way for users to start fresh with an object either by allowing them to easily wipe out the previous results or allowing them to name the file of the resulting RV curve.

bjfultn commented 4 years ago

The default behavior is indeed to include all observations of a given target going back to the beginning of the bulk reduction. However, it should be possible to exclude observations from the final RV timeseries using the "deactivation" features described in the Advanced Usage tutorial (https://caltech-ipac.github.io/hiresprv/tutorials/Advanced_Usage.html).

cgelino commented 4 years ago

I saw the “deactivate” feature, but there are 800+ data points in the full RV curve and I don’t want to have to deactivate all of them. That’s why I’m suggesting features where I can either ignore everything done before or create a new set of reduced products, leaving the old ones as is.

Dr. Christopher R. Gelino | Scientist | Caltech/IPAC Mail Code 100-22 | Pasadena CA 91125 | Office: 626-395-5809

From: BJ Fulton notifications@github.com Reply-To: Caltech-IPAC/hiresprv reply@reply.github.com Date: Tuesday, October 20, 2020 at 9:34 AM To: Caltech-IPAC/hiresprv hiresprv@noreply.github.com Cc: "Gelino, Christopher" cgelino@ipac.caltech.edu, Author author@noreply.github.com Subject: Re: [Caltech-IPAC/hiresprv] RV curve file contains all results for an object (#39)

The default behavior is indeed to include all observations of a given target going back to the beginning of the bulk reduction. However, it should be possible to exclude observations from the final RV timeseries using the "deactivation" features described in the Advanced Usage tutorial (https://caltech-ipac.github.io/hiresprv/tutorials/Advanced_Usage.html).

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/Caltech-IPAC/hiresprv/issues/39#issuecomment-712989234, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AKWNP3QGWR6IAVLHD67IYM3SLW3XVANCNFSM4SWOUNFQ.

bjfultn commented 4 years ago

I looked into implementing a convenience function to activate/deactivate all pre-reduced data. However we don't have a column in the database that allows us to differentiate pre-reduced data from new data. In addition, we allow the users to reprocess any of the old data and we have no way of knowing whether that has taken place.

Scientifically I don't think there is much of a use case for deactivating all of the pre-reduced RVs. The more RVs that are in the final time series the more reliable they all get due to the statistical weighting and outlier rejection that happens in the "vanking" phase. If users want to analyze only their new data they should still include most/all of the pre-processed data in the reduction but then simply ignore the measurements in the final RV timeseries.