[SPARK-15177.1] [R] make SparkR model params and default values consistent with MLlib #13801
Conversation
|
cc: @shivaram |
| #' @note spark.kmeans since 2.0.0 | ||
| setMethod("spark.kmeans", signature(data = "SparkDataFrame", formula = "formula"), | ||
| function(data, formula, k, maxIter = 10, initMode = c("random", "k-means||")) { | ||
| function(data, formula, k = 2, maxIter = 20, initMode = c("k-means||", "random")) { |
There was a problem hiding this comment.
just to clarify - this initMode change wasn't present in #13023 -- Is this intended to match some Spark behavior ?
There was a problem hiding this comment.
Yes, the default initMode in MLlib is k-means|| instead of random. See https://github.com/apache/spark/blob/master/mllib/src/main/scala/org/apache/spark/ml/clustering/KMeans.scala#L263.
There was a problem hiding this comment.
Ah I see - change LGTM then
|
Changes look fine given what was a part of #13023 |
|
test this please |
|
Test build #60913 has finished for PR 13801 at commit
|
|
Test build #60915 has finished for PR 13801 at commit
|
|
Test build #60917 has finished for PR 13801 at commit
|
|
Merged into master and branch-2.0. |
…istent with MLlib ## What changes were proposed in this pull request? This PR is a subset of #13023 by yanboliang to make SparkR model param names and default values consistent with MLlib. I tried to avoid other changes from #13023 to keep this PR minimal. I will send a follow-up PR to improve the documentation. Main changes: * `spark.glm`: epsilon -> tol, maxit -> maxIter * `spark.kmeans`: default k -> 2, default maxIter -> 20, default initMode -> "k-means||" * `spark.naiveBayes`: laplace -> smoothing, default 1.0 ## How was this patch tested? Existing unit tests. Author: Xiangrui Meng <[email protected]> Closes #13801 from mengxr/SPARK-15177.1. (cherry picked from commit 4f83ca1) Signed-off-by: Xiangrui Meng <[email protected]>
3 participants
What changes were proposed in this pull request?
This PR is a subset of #13023 by @yanboliang to make SparkR model param names and default values consistent with MLlib. I tried to avoid other changes from #13023 to keep this PR minimal. I will send a follow-up PR to improve the documentation.
Main changes:
spark.glm: epsilon -> tol, maxit -> maxIterspark.kmeans: default k -> 2, default maxIter -> 20, default initMode -> "k-means||"spark.naiveBayes: laplace -> smoothing, default 1.0How was this patch tested?
Existing unit tests.