SparkR from Rstudio - gives Error in invokeJava(is

问题:

I am using RStudio.

After creating session if i try to create dataframe using R data it gives error.

Sys.setenv(SPARK_HOME = "E:/spark-2.0.0-bin-hadoop2.7/spark-2.0.0-bin-hadoop2.7")
Sys.setenv(HADOOP_HOME = "E:/winutils")
.libPaths(c(file.path(Sys.getenv("SPARK_HOME"), "R", "lib"), .libPaths()))
Sys.setenv('SPARKR_SUBMIT_ARGS'='"sparkr-shell"')

library(SparkR)

sparkR.session(sparkConfig = list(spark.sql.warehouse.dir="C:/Temp"))

localDF <- data.frame(name=c("John", "Smith", "Sarah"), age=c(19, 23, 18))
df <- createDataFrame(localDF)

ERROR :

Error in invokeJava(isStatic = TRUE, className, methodName, ...) : 
  java.lang.reflect.InvocationTargetException
    at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
    at sun.reflect.NativeConstructorAccessorImpl.newInstance(Unknown Source)
    at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(Unknown Source)
    at java.lang.reflect.Constructor.newInstance(Unknown Source)
    at org.apache.spark.sql.hive.client.IsolatedClientLoader.createClient(IsolatedClientLoader.scala:258)
    at org.apache.spark.sql.hive.HiveUtils$.newClientForMetadata(HiveUtils.scala:359)
    at org.apache.spark.sql.hive.HiveUtils$.newClientForMetadata(HiveUtils.scala:263)
    at org.apache.spark.sql.hive.HiveSharedState.metadataHive$lzycompute(HiveSharedState.scala:39)
    at org.apache.spark.sql.hive.HiveSharedState.metadataHive(HiveSharedState.scala:38)
    at org.apache.spark.sql.hive.HiveSharedState.externalCatalog$lzycompute(HiveSharedState.scala:46)
    at org.apache.spark.sql.hive.HiveSharedState.externalCatalog(HiveSharedState.scala:45)
    at org.a
>

TIA.

回答1:

If you have not used SparkR library but you're using Spark, I recommend 'sparklyr' library made by RStudio.

1. Install the preview version of RStudio.

2. Install the library:

   install.packages("devtools")
   devtools::install_github('rstudio/sparklyr')
   

3. Load library and install spark.

   library(sparklyr)
   spark_install('1.6.2')
   

You can see a vignette in http://spark.rstudio.com/

回答2:

All many thanks for your help.

1. I had to do was set hadoop_home path in PATH variables (winutils/bin). This should have your winutils.exe file. So when it creates metastore for hive default derby) it is able to call hive classes.
2. Also i had set hive support as False as i am not using it.

---

Sys.setenv(SPARK_HOME='E:/spark-2.0.0-bin-hadoop2.7/spark-2.0.0-bin-hadoop2.7',HADOOP_HOME='E:/winutils')

.libPaths(c(file.path(Sys.getenv('SPARK_HOME'), 'R', 'lib'),.libPaths()))

Sys.setenv('SPARKR_SUBMIT_ARGS'='"sparkr-shell"')

library(SparkR)
library(rJava)

sparkR.session(enableHiveSupport = FALSE,master = "local[*]", sparkConfig = list(spark.driver.memory = "1g",spark.sql.warehouse.dir="E:/winutils/bin/"))

df <- as.DataFrame(iris)

回答3:

These are the steps that I did in RStudio and it worked for me:

Sys.setenv(SPARK_HOME="C:\\spark-1.6.1-bin-hadoop2.6")
.libPaths(c(file.path(Sys.getenv("SPARK_HOME"), "R", "lib"), .libPaths()))

library(SparkR)
sc <- sparkR.init(master="local")
sqlContext <- sparkRSQL.init(sc)

localDF <- data.frame(name=c("John", "Smith", "Sarah"), age=c(19, 23, 18))
df <- createDataFrame(sqlContext, localDF)