do_rf: Collect all data from store, prepare data, train random...

In ugzvot/rOstluft.rmweather: Meteorological normalisation with rmweather with data from rOstluft stores

Description

Collect all data from store, prepare data, train random forest

Usage

do_rf(
  config,
  store,
  description,
  n_trees = NULL,
  n_samples = NULL,
  prepare_data_only = TRUE,
  output_dir = NULL
)

Arguments

config	in list form from yaml::read_yaml()
store	containing all the data
description	of the model. Will be saved in statistics.yml
n_trees	Priority order is: argument n_trees > config$n_trees > default 500
n_samples	Priority order is: argument n_samples > config$n_samples > default 50
prepare_data_only	only prepare the data and then return
output_dir	if not NULL all data will be saved under output_dir\config$site\xxhash64 (xxhash64 is calculated from the return value from rmweather::rmw_do_all() with digest::digest())

Value

list with prepared data or model

See Also

rmweather::rmw_do_all()

Examples

# TODO

ugzvot/rOstluft.rmweather documentation built on May 5, 2020, 9:27 a.m.