search

BEAST-Fitting / beast

Bayesian Extinction And Stellar Tool
http://beast.readthedocs.io
22 stars 34 forks source link

Units for output 1D PDFs and stats #222

Open lea-hagen opened 6 years ago

lea-hagen commented 6 years ago

For the issue of PDFs for negative fluxes (#120), we wanted to display log flux values, so we're outputting the PDF/stats results for [filter]_wd_bias in symlog values. The other fluxes are in log values. Other fit quantities are both log and linear values. After conversation with @karllark, we thought it might be better to just have everything saved as linear values to avoid confusion. Before making a decision, there should be more discussion (hence creating an Issue).

galaxyumi commented 4 years ago

I also want to make sure that we computed the expectation values for fit parameters in log (e.g., log T, log age, etc.) in a linear space before saving as log values.

karllark commented 4 years ago

Good point. Are those the only two that we want to check we are doing the computation in linear units? Or are there more?