README.md

In signaux-faibles/MLlogr:

MLlogr

Simple and pure R utility to track ML experiments.

Create custom tracker

To create a custom tracker, use the class GenerticTracker, and define the desired fields in the initialization function. See documentation of the field function to see details on the options offered when creating a field.

GenericTracker$new(
  database = "database",
  collection = "collection",
  fields = list(
    field("my_custom_field1"),
    field("my_custom_field2")
  )
)

If you plan to use the same tracker regularly, or you want a more customized tracker with additional attributes etc., you can also define a new object, inheriting from GenericTracker, and define the fields you want in the initialization.

CustomTracker <- R6::R6Class("CustomTracker", 
  inherit = GenericTracker, 
  public = list(
    initialize = function(
      database, 
      collection){

      super$initialize(
        database = database, 
        collection = collection, 
        fields = list(
          field("my_custom_field1"), 
          field("my_custom_field2")
        )
      )
    }  
  )
)

signaux-faibles/MLlogr documentation built on June 27, 2019, 1:20 p.m.