topic Aggregating days to month sum in QlikView

Aggregating days to month sum

Wed, 23 Jul 2014 11:50:08 GMT

Hello dear community, i have a strange problem while using group by for the first time, maybe someone can pint me to the mistake i made.

The goal is, to aggregate all daily values which represent amounts that can be added (so, f.e. not temperatures)

So far, i found out, that aggregate does not work in the script.

Why would someone implement something, that only works in objects on the gui ??

So, the alternative would be load with the "group by" option, which loads the data that should be aggregated through a resident load.

Qlikview responds with an syntax error to the short load scrip below.

but if i remove

"sum(MV_Data) as MW_MonData" and correct the ",", i get the following:

"Aggregation expressions required by GROUP BY clause"

Values_Aggr_Month:

noconcatenate LOAD

ID_SENS,

ID_VALUE, // water, heating, electricity etc.

UNIT, // m3, kwh, liter

DATE_month, // f.e. 201401

NR_Building,

// MV_day, // former number of the day, not used in this table

MV_Data, // numerical content for ID_VALUE

sum(MV_Data) as MW_MonData // whats wrong whith this one ?

RESIDENT DaiylValues

group by DATE_month, ID_SENS;

Any suggestions ? Thanks in advance.

Re: Aggregating days to month sum

Anonymous — Wed, 23 Jul 2014 12:01:09 GMT

You need to all fields to group by and remove MV_DATA field.

Values_Aggr_Month:

noconcatenate LOAD

ID_SENS,

ID_VALUE, // water, heating, electricity etc.

UNIT, // m3, kwh, liter

DATE_month, // f.e. 201401

NR_Building,

sum(MV_Data) as MW_MonData // whats wrong whith this one ?

RESIDENT DaiylValues

group by DATE_month, ID_SENS,ID_VALUE,UNIT,NR_Building;

Re: Aggregating days to month sum

mahesh_agrawal — Wed, 23 Jul 2014 12:09:27 GMT

Whenever we do aggregation using group by, we have to add all the columns, which are to be shown, in group by clause except the column on which aggregation is being done.

Load col1,clo2,clo3,sum(col4)

Resident

Tab1

group by

col1,col2,col3;