Aggregate many records not using aggregation function

I have data files which consist of samples from a datalogger, which creates one record every half hour (or similar) with temperature readings. I need to aggregate the 48 half hourly records for one day into a single record for the day with 48 temperature values. Is there a way to get an Aggregate transformer to do this, please? Or any other way?
TIA
Bob irving

Hi Bob,

I would recommend to use http://doc.cloveretl.com/documentation/ … lizer.html

* add “measurement_day” field into input record
* sort input data by “measurement_day” field
* set “Key” property of Denormalizer to “measurement_day” field (this ensures only records of same day are processed together)
* prepare output data so it contains field “temperatures” of type List (http://doc.cloveretl.com/documentation/ … ields.html)
* in Denormalizer transformation just fill data into list using CTL2 functions - see http://doc.cloveretl.com/documentation/ … -ctl2.html

Other possibility is to use String for “temperatures” instead of List and store data as comma separated values.

I hope this helps.