feat: hotstart

R/Graph.R

@@ -459,6 +459,10 @@ Graph = R6Class("Graph",
       graph_load_namespaces(self, "predict")
       graph_reduce(self, input, "predict", single_input)
     },
+    hotstart = function(input, single_input = TRUE) {
+      graph_load_namespaces(self, "train")
+      graph_reduce(self, input, "hotstart", single_input)
+    },
     help = function(help_type = getOption("help_type")) {
       parts = strsplit(self$man, split = "::", fixed = TRUE)[[1]]
       match.fun("help")(parts[[2]], package = parts[[1]], help_type = help_type)

R/GraphLearner.R

@@ -454,6 +454,7 @@ GraphLearner = R6Class("GraphLearner", inherit = Learner,
       }
 
       on.exit({self$graph$state = NULL})
+
       self$graph$train(task)
       state = self$graph$state
       class(state) = c("graph_learner_model", class(state))
@@ -466,6 +467,23 @@ GraphLearner = R6Class("GraphLearner", inherit = Learner,
       assert_list(prediction, types = "Prediction", len = 1,
         .var.name = sprintf("Prediction returned by Graph %s", self$id))
       prediction[[1]]
+    },
+
+    .hotstart = function(task) {
+      if (!is.null(get0("validate", self))) {
+        some_pipeops_validate = some(pos_with_property(self, "validation"), function(po) !is.null(po$validate))
+        if (!some_pipeops_validate) {
+          lg$warn("GraphLearner '%s' specifies a validation set, but none of its PipeOps use it.", self$id)
+        }
+      }
+
+      on.exit({self$graph$state = NULL})
+      # copy hotstart state to graph
+      self$graph$state = self$state$model
+      self$graph$hotstart(task)
+      state = self$graph$state
+      class(state) = c("graph_learner_model", class(state))
+      state
     }
   )
 )

R/PipeOp.R

@@ -348,6 +348,11 @@ PipeOp = R6Class("PipeOp",
       output = check_types(self, output, "output", "predict")
       output
     },
+    hotstart = function(input) {
+      # default for all pipops is to just train them
+      # pipeops that can do hotstarting should overload this method
+      self$train(input)
+    },
     help = function(help_type = getOption("help_type")) {
       parts = strsplit(self$man, split = "::", fixed = TRUE)[[1]]
       match.fun("help")(parts[[2]], package = parts[[1]], help_type = help_type)

R/PipeOpLearner.R

@@ -112,6 +112,18 @@ PipeOpLearner = R6Class("PipeOpLearner", inherit = PipeOp,
         output = data.table(name = "output", train = "NULL", predict = out_type),
         tags = "learner", packages = learner$packages, properties = properties
       )
+    },
+    hotstart = function(input) {
+      on.exit({private$.learner$state = NULL})
+      # copy state to learner
+      learner = private$.learner
+      learner$state = self$state
+
+      # train learner with hotstarting
+      train_result = mlr3:::learner_train(learner, task = input[[1]], train_row_ids = NULL, mode = "hotstart")
+      self$state = train_result$learner$state
+
+      list(NULL)
     }
   ),
   active = list(