subset the parameter output of a performance_measures object
subset_mods(perf_meas_OBJ, bst_mods = 5, train_params = FALSE)
| perf_meas_OBJ | a resulted object of the performance_measures function |
|---|---|
| bst_mods | the number of models to select (rows of each algorithm) |
| train_params | if the subset should be based on train or test output |
a list of lists
This function takes the result from the performance_measures function, the number of best performing grid-models and a boolean specifying if the train or test predictions should be taken into account when subseting the data.frames and it returns a list with the optimal parameters for each algorithm.
if (FALSE) { bst_m = subset_mods(perf_meas_OBJ = perf, bst_mods = 5, train_params = FALSE) bst_m }