Algorithm

Pipeline

Description

Sequentially apply a list of transforms and a final estimator.

Algorithm Workflow

Iterate transformer steps: fit_transform()

Final step: fit()

Implementation Details

Implemented in `pipeline.R`.

fit=function(X,y){
  data <- X
  for(i in seq_len(n_steps)){
    step <- steps[[i]]
    if(pipeline_is_transformer(step)){
       if(pipeline_has_method(step,"fit_transform")){
         data <- step$fit_transform(data)
       } # ...
    } else if(pipeline_is_estimator(step)){
       step$fit(data,y)
    }
  }
}

Use Cases

General purpose.