aboutsummaryrefslogtreecommitdiff
path: root/dev
diff options
context:
space:
mode:
authornoelsmith <mail@noelsmith.com>2015-08-27 23:59:30 -0700
committerJoseph K. Bradley <joseph@databricks.com>2015-08-27 23:59:30 -0700
commit7583681e6b0824d7eed471dc4d8fa0b2addf9ffc (patch)
treed8cfb483586f9aa024fc328aaf515048426b664e /dev
parent89b943438512fcfb239c268b43431397de46cbcf (diff)
downloadspark-7583681e6b0824d7eed471dc4d8fa0b2addf9ffc.tar.gz
spark-7583681e6b0824d7eed471dc4d8fa0b2addf9ffc.tar.bz2
spark-7583681e6b0824d7eed471dc4d8fa0b2addf9ffc.zip
[SPARK-10188] [PYSPARK] Pyspark CrossValidator with RMSE selects incorrect model
* Added isLargerBetter() method to Pyspark Evaluator to match the Scala version. * JavaEvaluator delegates isLargerBetter() to underlying Scala object. * Added check for isLargerBetter() in CrossValidator to determine whether to use argmin or argmax. * Added test cases for where smaller is better (RMSE) and larger is better (R-Squared). (This contribution is my original work and that I license the work to the project under Sparks' open source license) Author: noelsmith <mail@noelsmith.com> Closes #8399 from noel-smith/pyspark-rmse-xval-fix.
Diffstat (limited to 'dev')
0 files changed, 0 insertions, 0 deletions