Some (important!) systems are missing stellar `err1` and `err2` values that should exist.

A weird problem. Some systems like TRAPPIST-1 and LHS 1140 show stellar mass values and uncertainties on the NASA Exoplanet Archive website and online table, but in the pscompars table we download, no values of st_masserr1 and st_masserr2 are found. It looks as though the data are there inside the st_massstr column, but in a string that would need to be parse individually.

This is a problem because I had been using 0-valued uncertainties as a way of filtering out bad data.

My two proposed solutions:

I submitted a help desk ticket to the archive to see if they might be able to help.
In the meantime, I'll temporarily lift the "no 0 uncertainty" filter, but I'll leave this Issue open until we can resolve it more carefully.

Here's the ticket I submitted:

Hi Amazing Heroes of the Exoplanet Archive! I'm coming to you with a weird, tiny, specific oddness that I discovered playing with your tables and API. I hope you might please be able to help? When I downloaded the Planetary Systems Composite Parameters table using the API, I found there were some planetary systems for which the `st_masserr1` and `st_masserr2` (= the stellar mass uncertainties) were listed as 0.0, even though the online portal shows finite mass uncertainties and the `st_massstr` column contained a string that included uncertainties. I would love to be able to easily extract the stellar mass and stellar mass uncertainties for every system, and for the vast majority of systems it works to use `st_mass` and `st_masserr1`/`st_masserr2`, but just for a tiny subset, the numerical mass uncertainties seem to be missing. If I filter by `st_masserr1=0` in the web interface, I can see 32 systems. All of them have a "Stellar Mass" column that shows a central value and some uncertainties as a prettily formatted string, but all of them have 0 uncertainty in the numerical columns. I attach the CSV of that filtered table at the end of this ticket, although it doesn't show the masses with uncertainties column (which is presumably rendering `st_massstr`?). In an email, I can share a screenshot of what the table looks like in the web browser. I would love to be able to please get these uncertainties. Do I have to parse the `st_massstr` column (which would be particularly annoying because it has some symmetric and some asymmetric uncertainties), or is there perhaps something else we can do to get those numerical values populated in `st_masserr` and `st_masserr2`? I found this with mass, but a quick look seems to show it's also a problem at least for radius and luminosity too. Thanks for your attention, and please let me know if there's anything I can do to help! I cannot express how grateful I am for all of you maintaining this amazing service! Cheers, Zach Berta-Thompson

PSCompPars_2024.09.06_09.36.49.csv

zkbt / exoatlas

Some (important!) systems are missing stellar `err1` and `err2` values that should exist. #58