search

BaM-tools / BaM

BaM (Bayesian Modeling) is a framework aimed at estimating a model using a bayesian approach and using it for prediction, with a particular focus on uncertainty quantification.
Other
2 stars 0 forks source link

Estimation without observations #15

Open JeromeLeCoz opened 4 months ago

JeromeLeCoz commented 4 months ago

It would be useful to allow Bayesian inference with no observations (eg rating curve without any gauging).

benRenard commented 3 months ago

I just checked and it's already feasible, under the following conditions:

  1. the data file specified in the first line of Config_Data.txt has to exist (even if it's empty).
  2. nHeaders and nObs (lines 2 and 3 of Config_Data.txt) have to be consistent with the content of the data file. The easiest way to go is to set nHeaders=0 and nObs=0 and leave the data file empty. Alternatively you may set nHeaders=1 and nObs=0 if you prefer to still write headers in the data file (and no actual data after that).

Please check that it allows implementing what you want to implement @IvanHeriver and @JeromeLeCoz, and if so, please close the ticket.

JeromeLeCoz commented 3 months ago

Thanks, Ben. It seems that it allows implementing what I suggested in BaRatinAGE. Ivan will tell.