Mailing-List: contact commits-help@spark.apache.org; run by ezmlm
Precedence: bulk
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
From: yliang@apache.org
To: commits@spark.apache.org
Message-Id: <758eaff57ef84280acd499ab5df94b70@git.apache.org>
Subject: spark git commit: [SPARK-19155][ML] Make family case insensitive in
 GLM
Date: Mon, 23 Jan 2017 08:54:16 +0000 (UTC)
archived-at: Mon, 23 Jan 2017 08:54:18 -0000

Repository: spark
Updated Branches:
  refs/heads/branch-2.1 8daf10e3f -> 1e07a7192


[SPARK-19155][ML] Make family case insensitive in GLM

## What changes were proposed in this pull request?
This is a supplement to PR #16516 which did not make the value from `getFamily` case insensitive. Current tests of poisson/binomial glm with weight fail when specifying 'Poisson' or 'Binomial', because the calculation of `dispersion` and `pValue` checks the value of family retrieved from `getFamily`
```
model.getFamily == Binomial.name || model.getFamily == Poisson.name
```

## How was this patch tested?
Update existing tests for 'Poisson' and 'Binomial'.

yanboliang felixcheung imatiach-msft

Author: actuaryzhang <actuaryzhang10@gmail.com>

Closes #16675 from actuaryzhang/family.

(cherry picked from commit f067acefabebf04939d03a639a2aaa654e1bc8f9)
Signed-off-by: Yanbo Liang <ybliang8@gmail.com>


Project: http://git-wip-us.apache.org/repos/asf/spark/repo
Commit: http://git-wip-us.apache.org/repos/asf/spark/commit/1e07a719
Tree: http://git-wip-us.apache.org/repos/asf/spark/tree/1e07a719
Diff: http://git-wip-us.apache.org/repos/asf/spark/diff/1e07a719

Branch: refs/heads/branch-2.1
Commit: 1e07a71924ef1420c96a3a0a8cb5be2f3a830037
Parents: 8daf10e
Author: actuaryzhang <actuaryzhang10@gmail.com>
Authored: Mon Jan 23 00:53:44 2017 -0800
Committer: Yanbo Liang <ybliang8@gmail.com>
Committed: Mon Jan 23 00:54:08 2017 -0800

----------------------------------------------------------------------
 .../spark/ml/regression/GeneralizedLinearRegression.scala      | 6 ++++--
 .../spark/ml/regression/GeneralizedLinearRegressionSuite.scala | 4 ++--
 2 files changed, 6 insertions(+), 4 deletions(-)
----------------------------------------------------------------------


http://git-wip-us.apache.org/repos/asf/spark/blob/1e07a719/mllib/src/main/scala/org/apache/spark/ml/regression/GeneralizedLinearRegression.scala
----------------------------------------------------------------------
diff --git a/mllib/src/main/scala/org/apache/spark/ml/regression/GeneralizedLinearRegression.scala b/mllib/src/main/scala/org/apache/spark/ml/regression/GeneralizedLinearRegression.scala
index 1e7ba91..676be61 100644
--- a/mllib/src/main/scala/org/apache/spark/ml/regression/GeneralizedLinearRegression.scala
+++ b/mllib/src/main/scala/org/apache/spark/ml/regression/GeneralizedLinearRegression.scala
@@ -1027,7 +1027,8 @@ class GeneralizedLinearRegressionSummary private[regression] (
    */
   @Since("2.0.0")
   lazy val dispersion: Double = if (
-    model.getFamily == Binomial.name || model.getFamily == Poisson.name) {
+    model.getFamily.toLowerCase == Binomial.name ||
+      model.getFamily.toLowerCase == Poisson.name) {
     1.0
   } else {
     val rss = pearsonResiduals.agg(sum(pow(col("pearsonResiduals"), 2.0))).first().getDouble(0)
@@ -1130,7 +1131,8 @@ class GeneralizedLinearRegressionTrainingSummary private[regression] (
   @Since("2.0.0")
   lazy val pValues: Array[Double] = {
     if (isNormalSolver) {
-      if (model.getFamily == Binomial.name || model.getFamily == Poisson.name) {
+      if (model.getFamily.toLowerCase == Binomial.name ||
+        model.getFamily.toLowerCase == Poisson.name) {
         tValues.map { x => 2.0 * (1.0 - dist.Gaussian(0.0, 1.0).cdf(math.abs(x))) }
       } else {
         tValues.map { x =>

http://git-wip-us.apache.org/repos/asf/spark/blob/1e07a719/mllib/src/test/scala/org/apache/spark/ml/regression/GeneralizedLinearRegressionSuite.scala
----------------------------------------------------------------------
diff --git a/mllib/src/test/scala/org/apache/spark/ml/regression/GeneralizedLinearRegressionSuite.scala b/mllib/src/test/scala/org/apache/spark/ml/regression/GeneralizedLinearRegressionSuite.scala
index 415d426..95b443d 100644
--- a/mllib/src/test/scala/org/apache/spark/ml/regression/GeneralizedLinearRegressionSuite.scala
+++ b/mllib/src/test/scala/org/apache/spark/ml/regression/GeneralizedLinearRegressionSuite.scala
@@ -757,7 +757,7 @@ class GeneralizedLinearRegressionSuite
        0.5554219 -0.4034267  0.6567520 -0.2611382
      */
     val trainer = new GeneralizedLinearRegression()
-      .setFamily("binomial")
+      .setFamily("Binomial")
       .setWeightCol("weight")
       .setFitIntercept(false)
 
@@ -874,7 +874,7 @@ class GeneralizedLinearRegressionSuite
        -0.4378554  0.2189277  0.1459518 -0.1094638
      */
     val trainer = new GeneralizedLinearRegression()
-      .setFamily("poisson")
+      .setFamily("Poisson")
       .setWeightCol("weight")
       .setFitIntercept(true)
 

---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@spark.apache.org
For additional commands, e-mail: commits-help@spark.apache.org