rstudio / spark-install Goto Github PK

Using vanilla win 10 VM, python.org 3.5.3, and java jdk 1.8.

Attempting install with no versions specified (thus gets latest) gives an error "No module named 'win32gui'
Workaround is to pip install pywin32

Need to either pull this module in if required or find another way to send broadcast message for the environment variable change.

Run spark_install(spark_version = "1.6.2"), currently CDH4 is installed; however, latest Hadoop version should be preffered.

See: https://github.com/rstudio/spark-install/blob/master/R/R/versions.R#L121-L125

Hi,
I can't use any installation who need a internet connection. My only solution is to download my file before in an other computer. Is it possible to download sources and install it after in an other commputer ?

In advance thank you

If there is an existing HADOOP_HOME environment variable, the new installation modifies it and sets it to null

Python prints the tuple with parenthesis and quotes which messes up the hive_site.xml file.