search

ICB-DCM / pyPESTO

python Parameter EStimation TOolbox
https://pypesto.readthedocs.io
BSD 3-Clause "New" or "Revised" License
211 stars 47 forks source link

Obtain correct parameter scale from petab import in parameter names #262

Open elbaraim opened 4 years ago

elbaraim commented 4 years ago

When petab problem is imported in pypesto would be good to update parameter names according to the scale: lin, log or log10. In case being lin leave as it is. In case any of the others change during import to, e.g., log(parameter_name) or log_{10}(parameter_name)

JanHasenauer commented 4 years ago

and/or to adapt the scale of the visualisation. e.g. visualise a log-transformed values on a log scale

yannikschaelte commented 4 years ago

in that case, it might be best to have an additional field in the pypesto.Problem class "x_scales", which takes the scale for each parameter. One just has to be aware that unlike in petab, in pypesto we assume all values to be scaled already, e.g. the nominal values and bounds. For optimization this is best this way, however when defining priors one will have to be a bit careful.