PUB-1153: set a level on a factor col

R/h2o-package/R/Classes.R

@@ -1648,6 +1648,14 @@ h2o.sub <- function(pattern, replacement, x, ignore.case = FALSE) {
   return(res)
 }
 
+h2o.setLevel <- function(x, level) {
+  expr <- paste("setLevel(", paste(x@key, deparse(level), sep = ","), ")", sep = "")
+  res <- .h2o.__exec2(x@h2o, expr)
+  res <- .h2o.exec2(res$dest_key, h2o = x@h2o, res$dest_key)
+  res@logic <- FALSE
+  res
+}
+
 trim <- function(x) {
   if (!inherits(x, "H2OParsedData")) stop("x must be an H2OParsedData object")
   .h2o.__unop2("trim", x)

R/h2o-package/man/h2o.setLevel.Rd

@@ -0,0 +1,22 @@
+\name{h2o.setLevel}
+\alias{h2o.setLevel}
+\title{Pattern Replacement}
+\description{ \code{h2o.setLevel}, a method to set a factor column to one of the levels.}
+\usage{h2o.setLevel(x, level)}
+\arguments{
+  \item{x}{An \code{\linkS4class{H2OParsedData}} object with a single factor column.}
+  \item{level}{The level at which the column will be set.}
+  }
+\details{
+  Replace all other occurrences with `level` in a factor column.
+}
+\value{
+An object of class "H2OParsedData".
+}
+
+\examples{
+library(h2o)
+localH2O <- h2o.init(ip = "localhost", port = 54321, startH2O = TRUE)
+hex <- as.h2o(localH2O  , iris)
+hex$Species <- h2o.setLevel(hex$Species, "versicolor")
+}

src/main/java/water/exec/ASTOp.java

@@ -21,8 +21,6 @@
 import java.util.*;
 import java.util.concurrent.atomic.AtomicInteger;
 
-import static water.util.Utils.seq;
-
 /** Parse a generic R string and build an AST, in the context of an H2O Cloud
  *  @author cliffc@0xdata.com
  */
@@ -183,6 +181,7 @@ public abstract class ASTOp extends AST {
     putPrefix(new ASTToLower());
     putPrefix(new ASTToUpper());
     putPrefix(new ASTGSub());
+    putPrefix(new ASTSetLevel());
     putPrefix(new ASTStrSub());
     putPrefix(new ASTRevalue());
     putPrefix(new ASTWhich());
@@ -994,6 +993,36 @@ class ASTGSub extends ASTOp {
   }
 }
 
+class ASTSetLevel extends ASTOp {
+  ASTSetLevel() { super(new String[]{"setLevel", "x", "level"},
+          new Type[]{Type.ARY, Type.ARY, Type.STR},
+          OPF_PREFIX,
+          OPP_PREFIX, OPA_RIGHT); }
+  @Override String opStr() { return "setLevel"; }
+  @Override ASTOp make() { return new ASTSetLevel(); }
+  @Override void apply(Env env, int argcnt, ASTApply apply) {
+    final String level = env.popStr();
+    String skey = env.key();
+    Frame fr = env.popAry();
+    if (fr.numCols() != 1) throw new IllegalArgumentException("setLevel works on a single column at a time.");
+    String[] doms = fr.anyVec().domain().clone();
+    if( doms == null )
+      throw new IllegalArgumentException("Cannot set the level on a non-factor column!");
+    final int idx = Arrays.asList(doms).indexOf(level);
+    if (idx == -1)
+      throw new IllegalArgumentException("Did not find level `" + level + "` in the column.");
+
+    Frame fr2 = new MRTask2() {
+      @Override public void map(Chunk c, NewChunk nc) {
+        for (int i=0;i<c._len;++i)
+          nc.addNum(idx);
+      }
+    }.doAll(1, fr.anyVec()).outputFrame(null, fr.names(), fr.domains());
+    env.subRef(fr, skey);
+    env.poppush(1, fr2, null);
+  }
+}
+
 class ASTTrim extends ASTOp {
   ASTTrim() { super(new String[]{"trim","x"},
           new Type[]{Type.dblary(), Type.dblary()},