new_model: Create a new data set metadata file.
In jbryer/mldash: Evaluating Machine Learning Models Across Many Datasets

new_model

R Documentation

Create a new data set metadata file.

Description

Create a new data set metadata file.

Usage

new_model(
  filename,
  name,
  type,
  dir = "./inst/models",
  description = "Description of the dataset",
  train_fun = "function(formula, data) {\n\t# Run predictive model.\n\t}",
  predict_fun =
    "function(model, newdata) {\n\t# Return predicted probabilities for classification or nemeric values for regression.\n\t}",
  packages = "",
  note = "",
  open = interactive(),
  overwrite = FALSE
)

Arguments

`filename`	the name of the file that will be created. This will be the id for the model.
`name`	the name of the model.
`type`	type of model, either regression (quantitative dependent variables), or classification (for qualitative dependent variables).
`dir`	the directory to store the model file.
`description`	brief description of the model.
`train_fun`	function that will return a trained model. Must have two parameters: `formula` and `data`.
`predict_fun`	function that will return the predicted probabilities for classification models and predicted Y-values for regression models.
`packages`	comma separated list of any packages that need to be loaded prior to running either `train_fun` or `predict_fun`.
`note`	any additional descriptions about the model.
`open`	whether to open the file for editing after it has been created.
`overwrite`	whether to overwrite the model file if it already exists.