makeTask: make a single mlr regression task for a single set of...

In pokyah/agrometeoR: Spatial interpolation for Agromet project @ CRA-W

Description

make a single mlr regression task for a single set of hourly/daily records

Usage

makeTask(dataset, drop = NULL, target)

Arguments

dataset	a dataframe containing an hourly/daily set of records you want to transform to a mlr task
drop	a character vector specifying the explanatory variables you want to drop.
target	a charachter specifying the name of the target variable

Value

A 2 elements named list

• snitch : a boolean. Is TRUE if function has provided the expected result. Is FALSE is function throws an error

• output : a named list which elements are :

  • value : a list which elements are of classes mlr::RegrTask()

  • condition : a character specifying the condition encountered by the function: success, warning, or error.

  • message : a character specifying the message relative to the condition.

  • stations : a named list which elements are

    • used : an integer vector of the sids of the used stations for the task creation

    • ignored an integer vector of the sids of the ignored stations for the task creation

Author(s)

Thomas Goossens

Examples

## Not run: 
# load magrittr for pipe use : %>%
library(magrittr)

# create the dataset
myDataset = makeDataset(
  dfrom = "2017-03-04T15:00:00Z",
  dto = "2017-03-04T18:00:00Z",
  sensor = "tsa")

# extract the list of hourly sets of records
myDataset = myDataset$output$value

# create the tasks
myTasks = purrr::map(myDataset, makeTask, target = "tsa")

# extract the tasks from the outputs
myTasks = myTasks %>% purrr::modify_depth(1, ~.$"output"$"value"$"task")

# show the first task information
myTasks[[1]]

# extract the lists of stations kept for each task
stations_kept = myTasks %>% purrr::modify_depth(1, ~.$"output"$"value"$"stations")

# show the stations kept for the first task
stations_kept[[1]]

## End(Not run)

pokyah/agrometeoR documentation built on May 26, 2019, 7 p.m.