Plot cross-validated results for the ML model and the trait under study.

Source: R/plot_results_cv.R

Prediction accuracy is always computed on an environmental basis (i.e. the correlations between the observed and predicted values are calculated within the same environment, no matter what is the cross-validation scheme used).

plot_results_cv(
  fitting_all_splits,
  trait,
  info_environments,
  cv_type,
  cv0_type,
  path_folder,
  nb_folds_cv1,
  repeats_cv1,
  nb_folds_cv2,
  repeats_cv2
)

Arguments

fitting_all_splits

results obtained using the function fit_cv_split() for all train/test partitions.
trait

character used in the predict_trait_MET_cv() function called.
info_environments

data.frame used in the predict_trait_MET_cv(). Typically METData$info_environments.
cv_type

character CV type used in the predict_trait_MET_cv() function called.
cv0_type

character For CV0 type, different possibilities are: "leave-one-environment-out", "leave-one-site-out", "leave-one-location-out".
path_folder

character Path where plots should be saved.
nb_folds_cv1

numeric Number of folds used in the CV1 scheme.
repeats_cv1

numeric Number of repeats in the CV1 scheme.
nb_folds_cv2

numeric Number of folds used in the CV2 scheme.
repeats_cv2

numeric Number of repeats in the CV2 scheme.

Value

Plots are directly saved in the path_folder.

Author

Cathy C. Westhues cathy.jubin@uni-goettingen.de