create_model: Train an RFSLAM model

In dshenker/RFSLAM: What the Package Does (One Line, Title Case)

Train an RFSLAM model

Description

create_model trains a RFSLAM model using the given user parameter values

Usage

create_model(
  modeling_df,
  target,
  id_col,
  risk_time_col,
  patient_count_col = "int.n",
  time_col,
  ntree = 100,
  nodedepth = NULL,
  nsplit = 10
)

Arguments

modeling_df	dataframe containing data for modeling
target	name of the target variable column
id_col	name of the column with the patient id's
risk_time_col	name of the column with the risk time values
patient_count_col	name of the column with the patient counts
ntree	number of trees for random forest
nodedepth	the node depth for random forest
nsplit	the nsplit parameter for random forest

Value

the RF-SLAM model, the predicted risk values (uncalibrated) together in a list that you can index into using the model and preds arguments

dshenker/RFSLAM documentation built on Sept. 18, 2022, 3:26 a.m.