A function to train and predict crossvalidation folds
and train one full model and predict any external validation data.
This function is primarily used internally but can be used when
running workflows interactively (see vignette basic zoon useage
)
RunModels(df, modelFunction, paras, workEnv)
Arguments
- df
- Dataframe from process module. Should contain columns
value, type, lon, lat
and fold, a number indicating which cross validation set a datapoint is in.
If fold is 0 then this is considered external validation data
If all data is 0 or 1, then no cross validation is run.
- modelFunction
- String giving the name of the model function which is
in turn the
name of the module.
- paras
- All other parameters that should be passed to the model function.
i.e. model[[1]]$paras
- workEnv
- The environment name of the workflow call environment.
Value
A list of length 2 containing the model trained on all data and a
data.frame which contains value, type, fold, lon, lat, predictions and
then all environmental variables.