Skip to content

Fix LALInference Gaussian Mixture Model prior for multiple values

Matthew Pitkin requested to merge matthew-pitkin/lalsuite:gmm_prior_varnames into master

Description

The Gaussian Mixture Model (GMM) prior in LALInferencePrior.c contains a bug that means that if there is more than one parameter being used, and values are drawn outside the prior ranges, the values get stuck and never change. This MR fixes this by making sure that previous values get removed if the value is outside the bounds. It also fixes another minor bug in the setting of a string variable.

The MR also makes a minor change to the ppe_likelihood code used when setting up a GMM from a prior file. It now check if the lower and upper bounds are finite independently of each other, rather than requiring both to be finite.

Notes: I have made the VARNAME_MAX macro in LALInference.h bigger, changing it from 40 to 50. This is to accommodate more GMM variables. If this might be problematic for any other LALInference code I can remove this.

API Changes and Justification

Backwards Compatible Changes

  • This change introduces no API changes
  • This change adds new API calls

Backwards Incompatible Changes

  • This change modifies an existing API
  • This change removes an existing API

If any of the Backwards Incompatible check boxes are ticked please provide a justification why this change is necessary and why it needs to be done in a backwards incompatible way.

Review Status

I have verified that the code compiles and works for a pulsar parameter estimation example. As this touches both LALInference and lalapps/pulsar code cc @vivien, @john-veitch, @david-keitel, @karl-wette.

Merge request reports