Skip to main content

mlr3 book

Quickstart

If you haven’t installed mlr3 already, do so now:

As a 30-second introductory example, we will train a decision tree model on the first 120 rows of iris data set and make predictions on the final 30, measuring the accuracy of the trained model.

library("mlr3")
task = tsk("iris")
learner = lrn("classif.rpart")

# train a model of this learner for a subset of the task
learner$train(task, row_ids = 1:120)
# this is what the decision tree looks like
learner$model
## n= 120 
## 
## node), split, n, loss, yval, (yprob)
##       * denotes terminal node
## 
## 1) root 120 70 setosa (0.41667 0.41667 0.16667)  
##   2) Petal.Length< 2.45 50  0 setosa (1.00000 0.00000 0.00000) *
##   3) Petal.Length>=2.45 70 20 versicolor (0.00000 0.71429 0.28571)  
##     6) Petal.Length< 4.95 49  1 versicolor (0.00000 0.97959 0.02041) *
##     7) Petal.Length>=4.95 21  2 virginica (0.00000 0.09524 0.90476) *
predictions = learner$predict(task, row_ids = 121:150)
predictions
## <PredictionClassif> for 30 observations:
##     row_ids     truth   response
##         121 virginica  virginica
##         122 virginica versicolor
##         123 virginica  virginica
## ---                             
##         148 virginica  virginica
##         149 virginica  virginica
##         150 virginica  virginica
# accuracy of our model on the test set of the final 30 rows
predictions$score(msr("classif.acc"))
## classif.acc 
##      0.8333

More examples can be found in the mlr3gallery, a collection of use cases and examples.

We highly recommend to keep some of our cheatsheets handy while learning mlr3.