Add a model for bagged neural networks

NAMESPACE

@@ -92,6 +92,7 @@ S3method(type_sum,model_fit)
 S3method(type_sum,model_spec)
 S3method(update,C5_rules)
 S3method(update,bag_mars)
+S3method(update,bag_mlp)
 S3method(update,bag_tree)
 S3method(update,bart)
 S3method(update,boost_tree)
@@ -146,6 +147,7 @@ export(augment)
 export(auto_ml)
 export(autoplot)
 export(bag_mars)
+export(bag_mlp)
 export(bag_tree)
 export(bart)
 export(bartMachine_interval_calc)

R/bag_mlp.R

@@ -0,0 +1,77 @@
+#' Ensembles of neural networks
+#'
+#' @description
+#'
+#' `bag_mlp()` defines an ensemble of single layer, feed-forward neural networks.
+#' This function can fit classification and regression models.
+#'
+#' \Sexpr[stage=render,results=rd]{parsnip:::make_engine_list("bag_mlp")}
+#'
+#' More information on how \pkg{parsnip} is used for modeling is at
+#' \url{https://www.tidymodels.org/}.
+#'
+#' @inheritParams mlp
+#'
+#' @template spec-details
+#'
+#' @template spec-references
+#'
+#' @seealso \Sexpr[stage=render,results=rd]{parsnip:::make_seealso_list("bag_mlp")}
+#' @export
+bag_mlp <-
+  function(mode = "unknown",
+           hidden_units = NULL,
+           penalty = NULL,
+           epochs = NULL,
+           engine = "nnet") {
+    args <- list(
+      hidden_units   = enquo(hidden_units),
+      penalty  = enquo(penalty),
+      epochs  = enquo(epochs)
+    )
+
+    new_model_spec(
+      "bag_mlp",
+      args = args,
+      eng_args = NULL,
+      mode = mode,
+      user_specified_mode = !missing(mode),
+      method = NULL,
+      engine = engine,
+      user_specified_engine = !missing(engine)
+    )
+  }
+
+# ------------------------------------------------------------------------------
+
+#' @method update bag_mlp
+#' @rdname parsnip_update
+#' @inheritParams mars
+#' @export
+update.bag_mlp <-
+  function(object,
+           parameters = NULL,
+           hidden_units = NULL, penalty = NULL, epochs = NULL,
+           fresh = FALSE, ...) {
+
+    args <- list(
+      hidden_units   = enquo(hidden_units),
+      penalty  = enquo(penalty),
+      epochs  = enquo(epochs)
+    )
+
+    update_spec(
+      object = object,
+      parameters = parameters,
+      args_enquo_list = args,
+      fresh = fresh,
+      cls = "bag_mlp",
+      ...
+    )
+  }
+
+# ------------------------------------------------------------------------------
+
+set_new_model("bag_mlp")
+set_model_mode("bag_mlp", "classification")
+set_model_mode("bag_mlp", "regression")

R/bag_mlp_nnet.R

@@ -0,0 +1,12 @@
+#' Bagged neural networks via nnet
+#'
+#' [baguette::bagger()] creates a collection of neural networks forming an
+#' ensemble. All trees in the ensemble are combined to produce a final prediction.
+#'
+#' @includeRmd man/rmd/bag_mlp_nnet.md details
+#'
+#' @name details_bag_mlp_nnet
+#' @keywords internal
+NULL
+
+# See inst/README-DOCS.md for a description of how these files are processed

R/print.R

@@ -38,6 +38,7 @@ model_descs <- tibble::tribble(
   ~cls,                   ~desc,
   "auto_ml",              "Automatic Machine Learning",
   "bag_mars",             "Bagged MARS",
+  "bag_mlp",              "Bagged Neural Network",
   "bag_tree",             "Bagged Decision Tree",
   "bart",                 "BART",
   "boost_tree",           "Boosted Tree",

_pkgdown.yml

@@ -35,6 +35,7 @@ reference:
     contents:
       - auto_ml
       - bag_mars
+      - bag_mlp
       - bag_tree
       - bart
       - boost_tree

inst/models.tsv

@@ -3,6 +3,8 @@
 "auto_ml"	"regression"	"h2o"	"agua"
 "bag_mars"	"classification"	"earth"	"baguette"
 "bag_mars"	"regression"	"earth"	"baguette"
+"bag_mlp"	"classification"	"nnet"	"baguette"
+"bag_mlp"	"regression"	"nnet"	"baguette"
 "bag_tree"	"censored regression"	"rpart"	"censored"
 "bag_tree"	"classification"	"C5.0"	"baguette"
 "bag_tree"	"classification"	"rpart"	"baguette"
@@ -12,10 +14,12 @@
 "boost_tree"	"censored regression"	"mboost"	"censored"
 "boost_tree"	"classification"	"C5.0"	NA
 "boost_tree"	"classification"	"h2o"	"agua"
+"boost_tree"	"classification"	"h2o_gbm"	"agua"
 "boost_tree"	"classification"	"lightgbm"	"bonsai"
 "boost_tree"	"classification"	"spark"	NA
 "boost_tree"	"classification"	"xgboost"	NA
 "boost_tree"	"regression"	"h2o"	"agua"
+"boost_tree"	"regression"	"h2o_gbm"	"agua"
 "boost_tree"	"regression"	"lightgbm"	"bonsai"
 "boost_tree"	"regression"	"spark"	NA
 "boost_tree"	"regression"	"xgboost"	NA
@@ -43,6 +47,7 @@
 "linear_reg"	"regression"	"brulee"	NA
 "linear_reg"	"regression"	"gee"	"multilevelmod"
 "linear_reg"	"regression"	"glm"	NA
+"linear_reg"	"regression"	"glmer"	"multilevelmod"
 "linear_reg"	"regression"	"glmnet"	NA
 "linear_reg"	"regression"	"gls"	"multilevelmod"
 "linear_reg"	"regression"	"h2o"	"agua"