[ https://issues.apache.org/jira/browse/SPARK-11205?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jeff Zhang updated SPARK-11205: ------------------------------- Description: When I use DataFrame#explain(), I found the output is a little different from scala API. Here's one example. {noformat} == Physical Plan == // this line is removed in pyspark API Scan JSONRelation[file:/Users/hadoop/github/spark/examples/src/main/resources/people.json][age#0L,name#1] {noformat} After looking at the code, I found that pyspark will print the output by itself rather than delegate it to spark-sql. This cause the difference between scala api and python api. I think both python api and scala api try to print it to standard out, so the python api can be delegated to scala api. Here's some api I found that can be delegated to scala api directly: * printSchema() * explain() * show() was: When I use DataFrame#explain(), I found the output is a little different from scala API. Here's one example. {noformat} == Physical Plan == // this line is removed in pyspark API Scan JSONRelation[file:/Users/hadoop/github/spark/examples/src/main/resources/people.json][age#0L,name#1] {noformat} After looking at the code, I found that pyspark will print the output by itself rather than delegate it to spark-sql. This cause the difference between scala api and python api. I think both python api and scala api try to print it to standard out, so the python api can be deleted to scala api. Here's some api I found that can be delegated to scala api directly: * printSchema() * explain() * show() > Delegate to scala DataFrame API rather than print in python > ----------------------------------------------------------- > > Key: SPARK-11205 > URL: https://issues.apache.org/jira/browse/SPARK-11205 > Project: Spark > Issue Type: Improvement > Components: PySpark > Affects Versions: 1.5.1 > Reporter: Jeff Zhang > Priority: Minor > > When I use DataFrame#explain(), I found the output is a little different from > scala API. Here's one example. > {noformat} > == Physical Plan == // this line is removed in pyspark API > Scan > JSONRelation[file:/Users/hadoop/github/spark/examples/src/main/resources/people.json][age#0L,name#1] > {noformat} > After looking at the code, I found that pyspark will print the output by > itself rather than delegate it to spark-sql. This cause the difference > between scala api and python api. I think both python api and scala api try > to print it to standard out, so the python api can be delegated to scala api. > Here's some api I found that can be delegated to scala api directly: > * printSchema() > * explain() > * show() -- This message was sent by Atlassian JIRA (v6.3.4#6332) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org