bind_gaugepreds: Bind gauge predictions

In NaiaraLopezRojo/globalIRmap: What the Package Does (One Line, Title Case)

Description

Compile predictions from multiple models. This function was used in the case when multiple models were developed for different (possibly overlapping) subsets of gauging stations.

Usage

bind_gaugepreds(in_gpredsdt, interthresh)

Arguments

in_gpredsdt	either a single data.table of predictions for gauges (output from make_gaugepreds) or a list of data.tables of model predictions to bind.
interthresh	either a single value or a data.table of values of the probability threshold to assign predicted flow intermittence classes to gauging stations (based on the predicted probability of flow intermittence).

Details

when several predictions are provided for the same gauging station (as was the case for this study for gauges with a long-term mean annual flow between 1 and 10 m3/s), the average predicted probability of flow intermittence was computed.

Value

data.table compiling all predictions with new columns of the predicted categorical flow intermittence class ("IRpredcat_') and flow intermittence prediction residuals (IPR; named preduncert_). The column suffixes reflect whether the predictions are based on the final model training or on cross-validation results (see make_gaugepreds).

NaiaraLopezRojo/globalIRmap documentation built on Dec. 17, 2021, 5:19 a.m.