Tasks

Tasks shipped with the mlr3 ecosystem.

The mlr3 packages also ship with some data sets, readily usable as Task objects. The goal of these tasks is to quickly demonstrate the capabilities of the packages.

Table

Example Usage

Create a classification task from the data set in the palmerpenguins package.

library("mlr3verse")

# create a task
task = tsk("breast_cancer")
task
<TaskClassif:breast_cancer> (683 x 10): Wisconsin Breast Cancer
* Target: class
* Properties: twoclass
* Features (9):
  - ord (9): bare_nuclei, bl_cromatin, cell_shape, cell_size,
    cl_thickness, epith_c_size, marg_adhesion, mitoses,
    normal_nucleoli
# get the dimensions
c(task$nrow, task$ncol)
[1] 683  10
# check for missing values
task$missings()
          class     bare_nuclei     bl_cromatin      cell_shape 
              0               0               0               0 
      cell_size    cl_thickness    epith_c_size   marg_adhesion 
              0               0               0               0 
        mitoses normal_nucleoli 
              0               0 
# plot class frequencies
autoplot(task)