search

California-Planet-Search / radvel

General Toolkit for Modeling Radial Velocity Data
http://radvel.readthedocs.io
MIT License
58 stars 53 forks source link

Gaussian Process Fitting - different Ks on plots? #320

Closed andreeaih closed 4 years ago

andreeaih commented 4 years ago

Hi, I am using the Gaussian Process feature from radvel and I was curious why is it that the semi-amplitude is different in GPPlot.plot_multipanel from the one from the Corner plot. This happens in the tutorial that you have posted (https://github.com/California-Planet-Search/radvel/blob/master/docs/tutorials/GaussianProcess-tutorial.ipynb). I ran radvel on my computer with GPs on a different dataset and I get a 1m/s difference between the two plots. I am guessing gp posterior - "gppost"- is not updated after the gp analysis was performed. Maybe there is something that I do not understand and this is how is supposed to be but I figured I could ask. Radvel version 1.4.1.

bjfultn commented 4 years ago

I believe plot_multipanel shows the maximum a posteriori value, but the corner plot shows the median value from the posterior. Check the table of parameters in the PDF report (which shows both) to verify that is the case.

andreeaih commented 4 years ago

Yes, you are right. Thank you for the help.