zhengruifeng edited a comment on pull request #30013:
URL: https://github.com/apache/spark/pull/30013#issuecomment-706965136
test in commit 27eab00:
```
import scala.util.Random
import org.apache.spark.ml.linalg._
import org.apache.spark.ml.classification._
import org.apache.spark.ml.regression._
import org.apache.spark.sql.functions._
import org.apache.spark.storage.StorageLevel
val df = spark.read.option("numFeatures",
"2000").format("libsvm").load("/data1/Datasets/epsilon/epsilon_normalized.t").withColumn("aftcensor",
(col("label")+1)/2).withColumn("aftlabel",
(col("label")+2)/2).withColumn("label", (col("label")+1)/2).limit(100)
df.persist(StorageLevel.MEMORY_AND_DISK)
df.count
val vec = df.select("features").head.getAs[Vector](0)
val lor = new LogisticRegression().setMaxIter(1).setThreshold(0.1)
val lorm = lor.fit(df)
lorm.getThreshold
lorm.predict(vec)
```
results:
// master
scala> val lorm = lor.fit(df)
20/10/12 15:47:23 WARN LogisticRegressionModel: **_threshold=0.5,
_rawThreshold=0.0**
lorm: org.apache.spark.ml.classification.LogisticRegressionModel =
LogisticRegressionModel: uid=logreg_4c79066a563d, numClasses=2, numFeatures=2000
scala> **lorm.getThreshold**
res9: Double = **0.1**
scala> lorm.predict(vec)
20/10/12 15:47:29 WARN LogisticRegressionModel: **_threshold=0.5,
_rawThreshold=0.0**
res10: Double = 0.0
The `_threshold` and `_rawThreshold` here are incorrect.
// this PR
scala> lorm.predict(vec)
20/10/12 16:01:09 WARN LogisticRegressionModel: _threshold=0.1,
_rawThreshold=-2.197224577336219
res3: Double = 0.0
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
