search

j-faria / kima

Exoplanet detection in RVs with DNest4 and GPs
http://www.kima.science
MIT License
15 stars 10 forks source link

Problems with priors? #34

Closed j-faria closed 6 years ago

j-faria commented 6 years ago

Running kima from the main directory with the maximum number of levels set to 1, followed by

kima-checkpriors 0 --code "samples=np.random.uniform(-47.6, 74, size=nsamples)"

gives something like

kima_prior_possible_bug

shouldn't they be more similar?

eggplantbren commented 6 years ago

Do you mean that the counts of the black curve are more variable over bins than the blue one? Perhaps try with more samples - but this could easily happen since it's MCMC rather than independent sampling, especially if the 'save interval' is low when you do this.

On Fri, May 4, 2018 at 3:26 PM, João Faria notifications@github.com wrote:

Running kima from the main directory with the maximum number of levels set to 1, followed by

kima-checkpriors 0 --code "samples=np.random.uniform(-47.6, 74, size=nsamples)"

gives something like

[image: kima_prior_possible_bug] https://user-images.githubusercontent.com/1958991/39611951-b3d6504c-4f31-11e8-8583-b0d7f4d1b1f0.png

shouldn't they be more similar?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/j-faria/kima/issues/34, or mute the thread https://github.com/notifications/unsubscribe-auth/ABgVOmtO4cW5UjgFJp4ou7YNNHS3CyOtks5tu8pfgaJpZM4TyESg .

-- Dr Brendon J. Brewer Department of Statistics, The University of Auckland, New Zealand Ph: +64 27 500 1336 Web: http://www.stat.auckland.ac.nz/~brewer/

j-faria commented 6 years ago

Of course, you're right that's it. I did set the save interval to 1, but I wasn't expecting such a difference. With more samples it's less striking. Thanks!