Update MeanSquare to Quadratic in R API for DL. Throw a warning if us…

…er passes MeanSquare to loss indicating it is deprecated in favor of Quadratic.
h2oai · Sep 23, 2015 · ee3d12d · ee3d12d
1 parent 3ae6d1c
commit ee3d12d
Show file tree

Hide file tree

Showing 2 changed files with 10 additions and 5 deletions.
diff --git a/h2o-r/ensemble/h2oEnsemble-package/R/wrappers.R b/h2o-r/ensemble/h2oEnsemble-package/R/wrappers.R
@@ -105,7 +105,7 @@ h2o.deeplearning.wrapper <- function(x, y, training_frame, model_id = "",
                                      hidden_dropout_ratios, l1 = 0, l2 = 0, max_w2 = Inf,
                                      initial_weight_distribution = c("UniformAdaptive", "Uniform", "Normal"),
                                      initial_weight_scale = 1,
-                                     loss = c("Automatic", "CrossEntropy", "MeanSquare", "Absolute", "Huber"), 
+                                     loss = c("Automatic", "CrossEntropy", "Quadratic", "Absolute", "Huber"), 
                                      distribution = c("AUTO", "gaussian", "bernoulli", "multinomial", 
                                                       "poisson", "gamma", "tweedie", "laplace", "huber"),
                                      tweedie_power = 1.5, score_interval = 5,

diff --git a/h2o-r/h2o-package/R/deeplearning.R b/h2o-r/h2o-package/R/deeplearning.R
@@ -47,7 +47,7 @@
 #' @param max_w2 Constraint for squared sum of incoming weights per unit (e.g. Rectifier)
 #' @param initial_weight_distribution Can be "Uniform", "UniformAdaptive", or "Normal"
 #' @param initial_weight_scale Uniform: -value ... value, Normal: stddev
-#' @param loss Loss function: "Automatic", "CrossEntropy" (for classification only), "MeanSquare", "Absolute"
+#' @param loss Loss function: "Automatic", "CrossEntropy" (for classification only), "Quadratic", "Absolute"
 #'        (experimental) or "Huber" (experimental)
 #' @param distribution A \code{character} string. The distribution function of the response.
 #'        Must be "AUTO", "bernoulli", "multinomial", "poisson", "gamma", "tweedie",
@@ -144,7 +144,7 @@ h2o.deeplearning <- function(x, y, training_frame,
                              max_w2 = Inf,
                              initial_weight_distribution = c("UniformAdaptive", "Uniform", "Normal"),
                              initial_weight_scale = 1,
-                             loss = c("Automatic", "CrossEntropy", "MeanSquare", "Absolute", "Huber"),
+                             loss = c("Automatic", "CrossEntropy", "Quadratic", "Absolute", "Huber"),
                              distribution = c("AUTO","gaussian", "bernoulli", "multinomial", "poisson", "gamma", "tweedie", "laplace", "huber"),
                              tweedie_power = 1.5,
                              score_interval = 5,
@@ -266,8 +266,13 @@ h2o.deeplearning <- function(x, y, training_frame,
     parms$initial_weight_distribution <- initial_weight_distribution
   if(!missing(initial_weight_scale))
     parms$initial_weight_scale <- initial_weight_scale
-  if(!missing(loss))
-    parms$loss <- loss
+  if(!missing(loss)) {
+    if(loss == "MeanSquare") {
+      warn("Loss name 'MeanSquare' is deprecated; please use 'Quadratic' instead.")
+      parms$loss <- "Quadratic"
+    } else
+      parms$loss <- loss
+  }
   if (!missing(distribution))
     parms$distribution <- distribution
   if (!missing(tweedie_power))