search

bob-carpenter / diagnostic-testing

statistical models to analyze diagnostic tests
15 stars 8 forks source link

Values for `mu_spec` etc #3

Closed maxbiostat closed 4 years ago

maxbiostat commented 4 years ago

I looked through the paper and the repo and could not find which values you guys used for mu_spec, mu_sens, sigma_sens and sigma_spec. Would it be possible to know which values you guys used? Thanks and apologies if I missed something really obvious.

bob-carpenter commented 4 years ago

The sigma_sens and sigma_spec from the plots are the hyperpriors on the scale of variation of sensitivity and specificity on the log odds scale. mu_spec and mu_sens are fitted parameters.`

The actual values for everything is in the R script

maxbiostat commented 4 years ago

Sorry I'm drag, but in this program, mu_sens is passed as data and that's what confused me. Is it fair to say that in the newest iteration of the manuscript at least you don't fit the model in Equation (2)?

bob-carpenter commented 4 years ago

The first one that's actually fit is the one in appendix A.2.

This kind of thing's a lot clearer when the paper's generated directly from rmarkdown. I never got around to converting the straight LaTeX drafts, so sorry it's confusing.

All the data's there and this stuff's super fast to execute because the data's so small and the models so simple.

maxbiostat commented 4 years ago

I never got around to converting the straight LaTeX drafts Neither did I for any of my papers/tech reports. I know it's probably best to have one master document, but this can get cumbersome at times.

Thanks for clarifying. I'll close the issue now.