Github user jkbradley commented on a diff in the pull request:
https://github.com/apache/spark/pull/7538#discussion_r36110220
--- Diff:
mllib/src/main/scala/org/apache/spark/ml/classification/LogisticRegression.scala
---
@@ -407,6 +449,103 @@ private[classification] class MultiClassSummarizer
extends Serializable {
}
}
+@Experimental
+/**
+ * :: Experimental ::
+ * Logistic regression training results.
+ * @param predictions dataframe outputted by the model's `transform`
method.
+ * @param probabilityCol field in "predictions" which gives the calibrated
probability of
+ * each sample as a vector.
+ * @param labelCol field in "predictions" which gives the true label of
each sample.
+ * @param objectiveHistory objective function (scaled loss +
regularization) at each iteration.
+ */
+class LogisticRegressionTrainingSummary private[classification] (
+ predictions: DataFrame,
+ probabilityCol: String,
+ labelCol: String,
+ val objectiveHistory: Array[Double])
+ extends LogisticRegressionSummary(predictions, probabilityCol, labelCol)
{
+
+ /** Number of training iterations until termination */
+ val totalIterations = objectiveHistory.length
+
+}
+
+@Experimental
+/**
+ * :: Experimental ::
+ * Logistic regression results for a given model.
+ * @param predictions dataframe outputted by the model's `transform`
method.
+ * @param probabilityCol field in "predictions" which gives the calibrated
probability of
+ * each sample.
+ * @param labelCol field in "predictions" which gives the true label of
each sample.
+ */
+class LogisticRegressionSummary private[classification] (
--- End diff --
Before merging, I still think we need to create the abstraction to handle
both binary and multiclass classification summaries. If we commit this PR now,
then we will need to break the API in the future to support multiclass
classification. Instead, we should prepare for it now by providing the
abstraction + the binary summary in this PR. (It's fine not to have the
multiclass version in this PR.)
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
