hi,
this is probably a very dumb question. i am using cosmomc as a generic sampler and produced a few triangle plots. in the 2-D plots, there are both contours and shadings. and i understand the contours are the confidence limits while the shadings show the likelihood distribution of the sample. Presumably, these two should follow each other. but i do see sometimes they differ. what is causing this?
similarly for the 1 D plots, the solid and dotted curves sometimes differ a lot. why?
Thanks.
reading the triangle plots
-
- Posts: 1
- Joined: March 06 2010
- Affiliation: Physics Department, Columbia Univeristy
- Contact:
-
- Posts: 1941
- Joined: September 23 2004
- Affiliation: University of Sussex
- Contact:
Re: reading the triangle plots
See appendix C of astro-ph/0205436.