Wrapper to generate multi-response predictive models.

mrIMLpredicts(
  X,
  Y,
  model1,
  balance_data = "no",
  model = "regression",
  parallel = TRUE,
  transformY = "log",
  tune_grid_size = 10,
  k = 10,
  seed = sample.int(1e+08, 1)
)

Arguments

X

A dataframe represents predictor or feature data.

Y

A dataframe is a response variable data set (species, OTUs, SNPs etc).

balance_data

A character 'up', 'down' or 'no'.

tune_grid_size

A numeric sets the grid size for hyperparamter tuning. Larger grid sizes increase computational time.

k

A numeric sets the number of folds in the 10-fold cross-validation. 10 is the default.

Model

1 A list can be any model from the tidy model package. See examples.

Details

This function produces yhats that used in all model characteristics for subsequent functions. This function fits separate classication models for each response variable in a dataset. Y (response variables) should be binary (0/1). Rows in X (features) have the same id (host/site/population) as Y. Class imblanace can be a real issue for classification analyses. Class imbalance can be addressed for each response variable using 'up' (upsampling using ROSE bootstrapping), 'down' (downsampling) or 'no' (no balancing of classes).