[SPARK-17157][SPARKR][WIP]: Add multiclass logistic regression SparkR Wrapper

R/pkg/NAMESPACE

@@ -43,7 +43,8 @@ exportMethods("glm",
               "spark.isoreg",
               "spark.gaussianMixture",
               "spark.als",
-              "spark.kstest")
+              "spark.kstest",
+              "spark.glmnet")
 
 # Job group lifecycle management methods
 export("setJobGroup",

R/pkg/R/generics.R

@@ -1304,6 +1304,11 @@ setGeneric("year", function(x) { standardGeneric("year") })
 #' @export
 setGeneric("spark.glm", function(data, formula, ...) { standardGeneric("spark.glm") })
 
+#' @rdname spark.glmnet
+#' @export
+setGeneric("spark.glmnet", function(data, formula, ...) { standardGeneric("spark.glmnet") })
+
+
 #' @param x,y For \code{glm}: logical values indicating whether the response vector
 #'          and model matrix used in the fitting process should be returned as
 #'          components of the returned value.

R/pkg/R/mllib.R

@@ -32,6 +32,13 @@
 #' @note GeneralizedLinearRegressionModel since 2.0.0
 setClass("GeneralizedLinearRegressionModel", representation(jobj = "jobj"))
 
+#' S4 class that represents a MultinomialLogisticRegressionModel
+#'
+#' @param jobj a Java object reference to the backing Scala MultinomialLogisticRegressionModel
+#' @export
+#' @note GeneralizedLinearRegressionModel since 2.1.0
+setClass("MultinomialLogisticRegressionModel", representation(jobj = "jobj"))
+
 #' S4 class that represents a NaiveBayesModel
 #'
 #' @param jobj a Java object reference to the backing Scala NaiveBayesWrapper
@@ -102,7 +109,7 @@ setClass("KSTest", representation(jobj = "jobj"))
 #' @rdname write.ml
 #' @name write.ml
 #' @export
-#' @seealso \link{spark.glm}, \link{glm},
+#' @seealso \link{spark.glm}, \link{glm}, \link{spark.glmnet},
 #' @seealso \link{spark.als}, \link{spark.gaussianMixture}, \link{spark.isoreg}, \link{spark.kmeans},
 #' @seealso \link{spark.lda}, \link{spark.mlp}, \link{spark.naiveBayes}, \link{spark.survreg}
 #' @seealso \link{read.ml}
@@ -115,7 +122,7 @@ NULL
 #' @rdname predict
 #' @name predict
 #' @export
-#' @seealso \link{spark.glm}, \link{glm},
+#' @seealso \link{spark.glm}, \link{glm}, \link{spark.glmnet},
 #' @seealso \link{spark.als}, \link{spark.gaussianMixture}, \link{spark.isoreg}, \link{spark.kmeans},
 #' @seealso \link{spark.mlp}, \link{spark.naiveBayes}, \link{spark.survreg}
 NULL
@@ -320,6 +327,54 @@ setMethod("predict", signature(object = "GeneralizedLinearRegressionModel"),
             predict_internal(object, newData)
           })
 
+setMethod("spark.glmnet", signature(data = "SparkDataFrame", formula = "formula"),
+          function(data, formula, regParam = 0.0, elasticNetParam = 0.0, tol = 1e-6, maxIter = 100,
+                   fitIntercept = TRUE, standardization = TRUE, thresholds = NULL, weightCol = NULL) {
+
+            formula <- paste0(deparse(formula), collapse = "")
+            if (is.null(weightCol)) {
+              weightCol <- ""
+            }
+
+            jobj <- callJStatic("org.apache.spark.ml.r.MultinomialLogisticRegressionWrapper",
+                                "fit", formula, data@sdf, as.numeric(regParam), as.numeric(elasticNetParam),
+                                tol, as.integer(maxIter), as.logical(fitIntercept),
+                                as.logical(standardization), as.array(thresholds), as.character(weightCol))
+            new("MultinomialLogisticRegressionModel", jobj = jobj)
+          })
+
+#  Predicted values based on a MultinomialLogisticRegression model
+
+#' @param object a fitted MultinomialLogisticRegressionModel
+#' @param newData SparkDataFrame for testing
+#' @return \code{predict} returns a SparkDataFrame containing predicted values
+#' @rdname spark.glmnet
+#' @aliases predict,MultinomialLogisticRegressionModel,SparkDataFrame-method
+#' @export
+#' @note predict(MultinomialLogisticRegressionModel) since 2.1.0
+setMethod("predict", signature(object = "MultinomialLogisticRegressionModel"),
+          function(object, newData) {
+            predict_internal(object, newData)
+          })
+
+#  Get the summary of a MultinomialLogisticRegression model
+
+#' @return \code{summary} returns the model's coefficients, intercepts and numClasses
+#' @rdname spark.glmnet
+#' @aliases summary,MultinomialLogisticRegression-method
+#' @export
+#' @note summary(MultinomialLogisticRegressionModel) since 2.1.0
+setMethod("summary", signature(object = "MultinomialLogisticRegressionModel"),
+          function(object) {
+            jobj <- object@jobj
+            coefficients <- callJMethod(jobj, "coefficients")
+            intercepts <- callJMethod(jobj, "intercepts")
+            numClasses <- callJMethod(jobj, "numClasses")
+            k <- callJMethod(jobj, "numFeatures")
+            coefficients <- t(matrix(coefficients, ncol = k))
+            list(coefficients = coefficients, intercepts = intercepts, numClasses = numClasses)
+          })
+
 # Makes predictions from a naive Bayes model or a model produced by spark.naiveBayes(),
 # similarly to R package e1071's predict.
 
@@ -826,6 +881,20 @@ setMethod("write.ml", signature(object = "GeneralizedLinearRegressionModel", pat
             write_internal(object, path, overwrite)
           })
 
+#  Saves the multinomial logistic regressionModel to the input path.
+
+#' @param path the directory where the model is saved.
+#' @param overwrite overwrites or not if the output path already exists. Default is FALSE
+#'                  which means throw exception if the output path exists.
+#'
+#' @rdname spark.glmnet
+#' @export
+#' @note write.ml(MultinomialLogisticRegressionModel, character) since 2.1.0
+setMethod("write.ml", signature(object = "MultinomialLogisticRegressionModel", path = "character"),
+          function(object, path, overwrite = FALSE) {
+            write_internal(object, path, overwrite)
+          })
+
 #  Save fitted MLlib model to the input path
 
 #' @param path the directory where the model is saved.
@@ -922,6 +991,8 @@ read.ml <- function(path) {
     new("GaussianMixtureModel", jobj = jobj)
   } else if (isInstanceOf(jobj, "org.apache.spark.ml.r.ALSWrapper")) {
     new("ALSModel", jobj = jobj)
+  } else if (isInstanceOf(jobj, "org.apache.spark.ml.r.MultinomialLogisticRegressionWrapper")) {
+    new("MultinomialLogisticRegressionModel", jobj = jobj)
   } else {
     stop("Unsupported model: ", jobj)
   }