Hi, this is my first post
I'm a beginner in Aster (nCluster v.6.0.1.00) and I am interested in using R scripts.
In the Aster document's I read in the section "Using the PARTITION BY Clause in Queries that Invoke R Programs" page 263 of the document "Aster_Database_User_Guide_0510" that "The PARTITION BY clause ensures that the rows that contain the same value for the partitioning (splitting) expression are processed by a single R program task"
I have a table with DISTRIBUTE BY HASH (ID) and i'm using the clausule of the STREAM function PARTITION BY ID.
I'm interested into execute the same R script for all the different values of the ID field, but i'm having troubles with this.
If the ID field have integer Type, all the ID values are processed into the same vworker in the same instance of the script (only one execution)
If the type is varchar, Aster makes the split using the filed but not one execution of the script for each value (i want obtain statistics for each value), instead this, Aster sends subsets of values of the field ID toward the script, sometimes one value and another times two values.
This ID values are linked to time series, and i want analyze each time-serie into independents executions of the script.
Why Aster send subsets and not the individual values of the ID field toward the script? How can i do this?
I'm very lost
Thank you in advance for reading this topic.