search

wwiecek / baggr

R package for Bayesian meta-analysis models, using Stan
GNU General Public License v3.0
46 stars 12 forks source link

standardisation and cross-validation need to work together #4

Open wwiecek opened 5 years ago

wwiecek commented 5 years ago

If user is using standardised data, they need to be standardised before splitting into testing and training datasets - currently this is not implemented.

wwiecek commented 4 years ago

This is now a problem that I will continue pushing forward until we have a case study where we are working with e.g. change from baseline. (TUP Graduation case study could work.) Reason for this is that standardisation is, in turn, related to whether we are standardising by control/baseline or the whole group. (See prepare_ma() for examples.)