The H2O prediction script is written in R. I use the following command to execute my code.
R CMD BATCH script.R _script.ou_t &
The output file doesn't show any error message but the data row count is always inconsistent.
I have been running my h2o prediction by importing the datafile which resides in hdfs.
The below command is used to import the file over.
DCPath <- h2o.importFolder(path=pathR,pattern="", parse=TRUE)
The file data count is actually 8.9 million rows but when the file is imported, it only reads about 3-4 millions rows.