support marshaling in GraphLearner

DESCRIPTION

@@ -93,7 +93,7 @@ Config/testthat/edition: 3
 Config/testthat/parallel: true
 NeedsCompilation: no
 Roxygen: list(markdown = TRUE, r6 = FALSE)
-RoxygenNote: 7.2.3
+RoxygenNote: 7.2.3.9000
 VignetteBuilder: knitr
 Collate:
  'Graph.R'

NAMESPACE

@@ -11,6 +11,7 @@ S3method(as_pipeop,Filter)
 S3method(as_pipeop,Learner)
 S3method(as_pipeop,PipeOp)
 S3method(as_pipeop,default)
+S3method(marshal_model,graph_learner_model)
 S3method(po,"NULL")
 S3method(po,Filter)
 S3method(po,Learner)
@@ -23,6 +24,7 @@ S3method(pos,list)
 S3method(predict,Graph)
 S3method(print,Multiplicity)
 S3method(print,Selector)
+S3method(unmarshal_model,graph_learner_model_marshalled)
 export("%>>!%")
 export("%>>%")
 export(Graph)
@@ -148,6 +150,7 @@ import(mlr3)
 import(mlr3misc)
 import(paradox)
 importFrom(R6,R6Class)
+importFrom(data.table,as.data.table)
 importFrom(digest,digest)
 importFrom(stats,setNames)
 importFrom(utils,bibentry)

R/Graph.R

@@ -63,6 +63,8 @@
 #' Whether to store intermediate results in the [`PipeOp`]'s `$.result` slot, mostly for debugging purposes. Default `FALSE`.
 #' * `man` :: `character(1)`\cr
 #' Identifying string of the help page that shows with `help()`.
+#' * `properties` :: `character()`\cr
+#' The properties of the `Graph` is the union of all the properties of its [`PipeOp`]s.
 #'
 #' @section Methods:
 #' * `ids(sorted = FALSE)` \cr
@@ -504,6 +506,10 @@ Graph = R6Class("Graph",
  } else {
  map(self$pipeops, "state")
  }
+ },
+ properties = function(rhs) {
+ assert_ro_binding(rhs)
+ sort(unique(unlist(map(self$pipeops, "properties"))))
  }
  ),
 

R/GraphLearner.R

@@ -98,11 +98,17 @@ GraphLearner = R6Class("GraphLearner", inherit = Learner,
  }
  assert_subset(task_type, mlr_reflections$task_types$type)
 
+ properties = mlr_reflections$learner_properties[[task_type]]
+
+ if ("marshal" %nin% graph$properties) {
+ propertiers = setdiff(properties, "marshal")
+ }
+
  super$initialize(id = id, task_type = task_type,
  feature_types = mlr_reflections$task_feature_types,
  predict_types = names(mlr_reflections$learner_predict_types[[task_type]]),
  packages = graph$packages,
- properties = mlr_reflections$learner_properties[[task_type]],
+ properties = properties,
  man = "mlr3pipelines::GraphLearner"
  )
 
@@ -132,9 +138,18 @@ GraphLearner = R6Class("GraphLearner", inherit = Learner,
  if (length(last_pipeop_id) == 0) stop("No Learner PipeOp found.")
  }
  learner_model$base_learner(recursive - 1)
+ },
+ marshal = function() {
+ learner_marshal(self)
+ },
+ unmarshal = function() {
+ learner_unmarshal(self)
  }
  ),
  active = list(
+ marshalled = function(rhs) {
+ learner_marshalled(self)
+ },
  hash = function() {
  digest(list(class(self), self$id, self$graph$hash, private$.predict_type,
  self$fallback$hash, self$parallel_predict), algo = "xxhash64")
@@ -189,6 +204,7 @@ 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))
  state
  },
  .predict = function(task) {
@@ -233,6 +249,20 @@ GraphLearner = R6Class("GraphLearner", inherit = Learner,
  )
 )
 
+#' @export
+marshal_model.graph_learner_model = function(model, ...) {
+ model = map(model, marshal_model)
+ class(model) = c("graph_learner_model_marshalled", "list_marshalled", "marshalled")
+ model
+}
+
+#' @export
+unmarshal_model.graph_learner_model_marshalled = function(model, ...) {
+ model = map(model, marshal_model)
+ class(model) = c("graph_learner_model", "list")
+ model
+}
+
 #' @export
 as_learner.Graph = function(x, clone = FALSE, ...) {
  GraphLearner$new(x, clone_graph = clone)

R/PipeOp.R

@@ -135,6 +135,8 @@
 #' and done, if requested, by the [`Graph`] backend itself; it should *not* be done explicitly by `private$.train()` or `private$.predict()`.
 #' * `man` :: `character(1)`\cr
 #' Identifying string of the help page that shows with `help()`.
+#' * `properties` :: `character()`\cr
+#' The properties that are this PipeOp has. See `mlr_reflections$pipeops$properties` for available values.
 #'
 #' @section Methods:
 #' * `train(input)`\cr
@@ -236,7 +238,7 @@ PipeOp = R6Class("PipeOp",
  .result = NULL,
  tags = NULL,
 
- initialize = function(id, param_set = ParamSet$new(), param_vals = list(), input, output, packages = character(0), tags = "abstract") {
+ initialize = function(id, param_set = ParamSet$new(), param_vals = list(), input, output, packages = character(0), tags = "abstract", properties = character(0)) {
  if (inherits(param_set, "ParamSet")) {
  private$.param_set = assert_param_set(param_set)
  private$.param_set_source = NULL
@@ -246,6 +248,7 @@ PipeOp = R6Class("PipeOp",
  }
  self$id = assert_string(id)
 
+ private$.properties = sort(assert_subset(properties, mlr_reflections$pipeops$properties))
  self$param_set$values = insert_named(self$param_set$values, param_vals)
  self$input = assert_connection_table(input)
  self$output = assert_connection_table(output)
@@ -411,6 +414,10 @@ PipeOp = R6Class("PipeOp",
  }
  }
  private$.label
+ },
+ properties = function(rhs) {
+ assert_ro_binding(rhs)
+ private$.properties
  }
  ),
 
@@ -429,6 +436,7 @@ PipeOp = R6Class("PipeOp",
  }
  value
  },
+ .properties = NULL,
  .train = function(input) stop("abstract"),
  .predict = function(input) stop("abstract"),
  .additional_phash_input = function() {

R/PipeOpLearner.R

@@ -89,10 +89,11 @@ PipeOpLearner = R6Class("PipeOpLearner", inherit = PipeOp,
  type = private$.learner$task_type
  task_type = mlr_reflections$task_types[type, mult = "first"]$task
  out_type = mlr_reflections$task_types[type, mult = "first"]$prediction
+ properties = if ("marshal" %in% learner$properties) "marshal" else character(0)
  super$initialize(id, param_set = alist(private$.learner$param_set), param_vals = param_vals,
  input = data.table(name = "input", train = task_type, predict = task_type),
  output = data.table(name = "output", train = "NULL", predict = out_type),
- tags = "learner", packages = learner$packages
+ tags = "learner", packages = learner$packages, properties = properties
  )
  }
  ),

R/PipeOpLearnerCV.R

@@ -122,6 +122,8 @@ PipeOpLearnerCV = R6Class("PipeOpLearnerCV",
  type = private$.learner$task_type
  task_type = mlr_reflections$task_types[type, mult = "first"]$task
 
+ properties =if ("marshal" %in% learner$properties) "marshal" else character(0)
+
  private$.crossval_param_set = ParamSet$new(params = list(
  ParamFct$new("method", levels = c("cv", "insample"), tags = c("train", "required")),
  ParamInt$new("folds", lower = 2L, upper = Inf, tags = c("train", "required")),
@@ -137,7 +139,7 @@ PipeOpLearnerCV = R6Class("PipeOpLearnerCV",
  # in PipeOp ParamSets.
  # private$.crossval_param_set$add_dep("folds", "method", CondEqual$new("cv")) # don't do this.
 
- super$initialize(id, alist(private$.crossval_param_set, private$.learner$param_set), param_vals = param_vals, can_subset_cols = TRUE, task_type = task_type, tags = c("learner", "ensemble"))
+ super$initialize(id, alist(private$.crossval_param_set, private$.learner$param_set), param_vals = param_vals, can_subset_cols = TRUE, task_type = task_type, tags = c("learner", "ensemble"), properties = properties)
  }
 
  ),

R/PipeOpTaskPreproc.R

@@ -169,7 +169,7 @@ PipeOpTaskPreproc = R6Class("PipeOpTaskPreproc",
 
  public = list(
  initialize = function(id, param_set = ParamSet$new(), param_vals = list(), can_subset_cols = TRUE,
- packages = character(0), task_type = "Task", tags = NULL, feature_types = mlr_reflections$task_feature_types) {
+ packages = character(0), task_type = "Task", tags = NULL, feature_types = mlr_reflections$task_feature_types, properties = character(0)) {
  if (can_subset_cols) {
  acp = ParamUty$new("affect_columns", custom_check = check_function_or_null, default = selector_all(), tags = "train")
  if (inherits(param_set, "ParamSet")) {
@@ -183,7 +183,8 @@ PipeOpTaskPreproc = R6Class("PipeOpTaskPreproc",
  super$initialize(id = id, param_set = param_set, param_vals = param_vals,
  input = data.table(name = "input", train = task_type, predict = task_type),
  output = data.table(name = "output", train = task_type, predict = task_type),
- packages = packages, tags = c(tags, "data transform")
+ packages = packages, tags = c(tags, "data transform"),
+ properties = properties
  )
  }
  ),

R/zzz.R

@@ -16,6 +16,8 @@ register_mlr3 = function() {
  c("abstract", "meta", "missings", "feature selection", "imbalanced data",
  "data transform", "target transform", "ensemble", "robustify", "learner", "encode",
  "multiplicity")))
+
+ x$pipeops$properties = "marshal"
 }
 
 .onLoad = function(libname, pkgname) { # nocov start

tests/testthat/test_GraphLearner.R

@@ -569,4 +569,15 @@ test_that("GraphLearner hashes", {
 
 })
 
+test_that("marshal", {
+ task = tsk("iris")
+ po_lily = as_pipeop(lrn("classif.lily"))
+ graph = as_graph(po_lily)
+ glrn = as_learner(graph)
+ expect_true("marshal" %in% glrn$properties)
 
+ # als checks that it is marshallable
+ expect_learner(glrn, task)
+
+ expect_false("marshal" %in% as_graph(lrn("regr.featureless"))$properties)
+})