search

commfish / southeast_pink_salmon_preseason

Preseason pink salmon forecasts.
0 stars 1 forks source link

Logging temp. in the regression model #6

Open fssem1 opened 4 years ago

fssem1 commented 4 years ago

Why is the temperature data logged?

Did this document come out yet? Murphy, J. M., E.A. Fergusson, A. Piston, A. Gray, and E. Farley. In Press. Growth and harvest forecast models for Southeast Alaska pink salmon. North Pacific Anadromous Fish Commisson Technical Report No. 15:xx-xx.

jmurphygithub commented 4 years ago

Not sure why or if it makes sense to log transform temperature.

The report is available online at NPAFC.... https://npafc.org/wp-content/uploads/technical-reports/Tech-Report-15-DOI/22_Murphy-et-al.pdf

jmurphygithub commented 4 years ago

I believe the decision to log transform temperature would be determined by model performance. If the model performs better without transforming temperature then I don't see why we would transform temperature.

fssem1 commented 4 years ago

For the 2020 forecast, the decision was made to log the temperature data. It looks like in past forecasts, the temperature data was not logged.