Update docstring for binary param

Author: holdenk

mllib/src/main/scala/org/apache/spark/ml/param/shared/sharedParams.scala

@@ -396,10 +396,10 @@ private[ml] trait HasSolver extends Params {
 private[ml] trait HasBinary extends Params {
 
   /**
-   * Param for If true, all non zero results are set to 1. This is useful for discrete probabilistic models that model binary events rather than integer counts. Default False..
+   * Param for If true, all non-zero counts (after any filters are applied) are set to 1. This is useful for discrete probabilistic models that model binary events rather than integer counts. Default False..
    * @group param
    */
-  final val binary: BooleanParam = new BooleanParam(this, "binary", "If true, all non zero results are set to 1. This is useful for discrete probabilistic models that model binary events rather than integer counts. Default False.")
+  final val binary: BooleanParam = new BooleanParam(this, "binary", "If true, all non-zero counts (after any filters are applied) are set to 1. This is useful for discrete probabilistic models that model binary events rather than integer counts. Default False.")
 
   setDefault(binary, false)
 

python/pyspark/ml/param/shared.py

@@ -585,10 +585,10 @@ def getVarianceCol(self):
 
 class HasBinary(Params):
     """
-    Mixin for param binary: If True, all non zero results are set to 1. This is useful for discrete probabilistic models that model binary events rather than integer counts. Default False.
+    Mixin for param binary: If True, all non-zero counts (after any filters are applied) are set to 1. This is useful for discrete probabilistic models that model binary events rather than integer counts. Default False.
     """
 
-    binary = Param(Params._dummy(), "binary", "If True, all non zero results are set to 1. This is useful for discrete probabilistic models that model binary events rather than integer counts. Default False.", typeConverter=TypeConverters.toBoolean)
+    binary = Param(Params._dummy(), "binary", "If True, all non-zero counts (after any filters are applied) are set to 1. This is useful for discrete probabilistic models that model binary events rather than integer counts. Default False.", typeConverter=TypeConverters.toBoolean)
 
     def __init__(self):
         super(HasBinary, self).__init__()