Skip to content

[SPARK-7390][SQL] Only merge other CovarianceCounter when its count is greater than zero #5931

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 1 commit into from
Closed

[SPARK-7390][SQL] Only merge other CovarianceCounter when its count is greater than zero #5931

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 12 additions & 10 deletions sql/core/src/main/scala/org/apache/spark/sql/execution/stat/StatFunctions.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,7 +38,7 @@ private[sql] object StatFunctions extends Logging {
var yAvg = 0.0 // the mean of all examples seen so far in col2
var Ck = 0.0 // the co-moment after k examples
var MkX = 0.0 // sum of squares of differences from the (current) mean for col1
var MkY = 0.0 // sum of squares of differences from the (current) mean for col1
var MkY = 0.0 // sum of squares of differences from the (current) mean for col2
var count = 0L // count of observed examples
// add an example to the calculation
def add(x: Double, y: Double): this.type = {
Expand All @@ -55,15 +55,17 @@ private[sql] object StatFunctions extends Logging {
// merge counters from other partitions. Formula can be found at:
// http://en.wikipedia.org/wiki/Algorithms_for_calculating_variance
def merge(other: CovarianceCounter): this.type = {
val totalCount = count + other.count
val deltaX = xAvg - other.xAvg
val deltaY = yAvg - other.yAvg
Ck += other.Ck + deltaX * deltaY * count / totalCount * other.count
xAvg = (xAvg * count + other.xAvg * other.count) / totalCount
yAvg = (yAvg * count + other.yAvg * other.count) / totalCount
MkX += other.MkX + deltaX * deltaX * count / totalCount * other.count
MkY += other.MkY + deltaY * deltaY * count / totalCount * other.count
count = totalCount
if (other.count > 0) {
val totalCount = count + other.count
val deltaX = xAvg - other.xAvg
val deltaY = yAvg - other.yAvg
Ck += other.Ck + deltaX * deltaY * count / totalCount * other.count
xAvg = (xAvg * count + other.xAvg * other.count) / totalCount
yAvg = (yAvg * count + other.yAvg * other.count) / totalCount
MkX += other.MkX + deltaX * deltaX * count / totalCount * other.count
MkY += other.MkY + deltaY * deltaY * count / totalCount * other.count
count = totalCount
}
this
}
// return the sample covariance for the observed examples
Expand Down