You should think of both the data and metadata folders and all of the composite XDF files stored in Hadoop in the 'data' folder
as really all of the information for a single XDF output file. In Hadoop we create a set of composite XDF files(you will see files
with name (*.xdfm) in the data folder. The metadata folder contains footer information for the XDF file such as number of variables, etc
that is needed for fast file access.
The reason we store XDF files in Hadoop this way is for fast file access and to take advantage of HDFS block sizes.
You will need to store the results from rxPredict() as a composite XDf file in Hadoop or as a composite CSV file in HDFS.
as really all of the information for a single XDF output file. In Hadoop we create a set of composite XDF files(you will see files
with name (*.xdfm) in the data folder. The metadata folder contains footer information for the XDF file such as number of variables, etc
that is needed for fast file access.
The reason we store XDF files in Hadoop this way is for fast file access and to take advantage of HDFS block sizes.
You will need to store the results from rxPredict() as a composite XDf file in Hadoop or as a composite CSV file in HDFS.