Dear Sergei,

The lhc.bf script requires a user input inflation factor.  What does this do, and how does one go about determining what value to use?

-Danny

Dear Danny,

The input inflation factor can be thought of a way to inflate your 'prior'. The reason for doing this is that if your model parameters are tightly correlated, the confidence intervals obtained by profile likelihood will be too small. The inflation factors are not rigorously defined -- they depend on the model in question and other many parameters. You can try a variety of settings (e.g. from 1.0 to 10.0) and see how this affects your sampling estimates.

Cheers,
Sergei