MSTI feature added, temp replaced by beta #56
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Model-Switch Thermodynamic Integration (MSTI) feature has been added, allowing users to sample across the posteriors of two models, instead of sampling between each model's prior and posterior, individually. Beta=0 is the posterior of model 0 and beta=1 is the posterior of model 1. A beta in between 0 and 1 is a mixture of the two models’ posterior distributions. Changes include the ability to give PTMCMC a tuple of prior and likelihood functions, a parameter of a tuple of lists of indices for each model’s parameters in MSTI used to sift through the combined set of parameters in p0, and a parameter to indicate the preferred ladder shape (only geometric or linear as of right now). When using MSTI, four additional columns are written in the chain file -- the log posterior and log likelihood of the two models being compared. Instances of 1/temperature have been replaced by beta. Changes include PTMCMC accepting beta max and beta min (while still accepting Tmin and Tmax), chain files named with beta instead of temperature, and a parameter to indicate whether the temperature naming convention should be use. Temperature scaling has been commented out since for hot chains it produces massive and frequently invalid jumps.
E501, E711, E262, E203, F821, E266
|
Looks good to me. @vhaasteren , @AaronDJohnson do you have any comments? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Model-Switch Thermodynamic Integration (MSTI) feature has been added, allowing users to sample across the posteriors of two models, instead of sampling between each model's prior and posterior, individually. Beta=0 is the posterior of model 0 and beta=1 is the posterior of model 1. A beta in between 0 and 1 is a mixture of the two models’ posterior distributions. Changes include the ability to give PTMCMC a tuple of prior and likelihood functions, a new parameter for a tuple of lists of indices for each model’s parameters used to sift through the combined set of parameters in p0, and a parameter to indicate the preferred ladder shape (only geometric or linear as of right now). When using MSTI, four additional columns are written in the chain file -- the log posterior and log likelihood of the two models being compared.
Instances of 1/temperature have been replaced by beta. Changes include PTMCMC accepting beta max and beta min (while still accepting Tmin and Tmax), chain files named with beta instead of temperature, and a parameter to indicate whether the temperature naming convention should be use instead.
Temperature scaling has been commented out since for hot chains it produces massive and frequently invalid jumps.