Hi everyone,
I´m new with SAP HANA and PAL Library. I have an analysis which consist in processing a table that have near 200 million of rows and tree columns. First is a consumer ID, second is a date and the third is double value. This table have data from 5.3 millions of consumers over 36 months. My analysis consist in execute a linear regression over this 36 values for each consumer (about 5.3 times). I need to hold only a regression betas and R2 for each consumer.
I´m trying to do this using HANA and Predictive Analysis (PAL). I´m able to execute linear regression for one consumer at time, I’m follow some examples in this forum and it´s works.
But there is some way to do this over all my data without using any LOOP pr CURSOR statement? Maybe using a CALL to PAL function inside an aggregation function?
Actually I’m doing this process using ORACLE aggregation function (REGR_SLOPE and REGR_R2).
Tks