Aggregator Stage

[dtatem]

OK,

I am using the aggregator stage and after aggregating the specified column, I want to the other columns to be outputted also, in the output link so I can carry them downstream. Any ideas?

thanks, dtatem

[ray.wurlod]

Same rules as SQL - every input column must either be grouped or have an aggregate function applied.

Think about it - fewer rows out than in, because of grouping. From which input rows would your passthrough columns be selected?

You could create a "fork join" design - split your data into two streams (Copy stage), run one stream through Aggregator, then bring both streams back together in an inner join in a Join stage.

[dtatem]

I forgot to mention something, after the aggregation is done on that input column, I want only one output showing me the Total sum of all rows after the aggregation, plus the other columns from the input data[no aggregation on these columns]

I have create a Groupkey for the aggregation plus the column which is to be aggregated. I guess I would have to join the aggregated value back to the other input rows so I can carry them downstream.

Thanks you all are a BIG help..

--dtatem

[JoshGeorge]

Yes you guessed it right

[vijayrc]

I forgot to mention something, after the aggregation is done on that input column, I want only one output showing me the Total sum of all rows after the aggregation, plus the other columns from the input data[no aggregation on these columns]

I have create a Groupkey for the aggregation plus the column which is to be aggregated. I guess I would have to join the aggregated value back to the other input rows so I can carry them downstream.

Thanks you all are a BIG help..

--dtatem

Other option would be to do the 'aggregation' using a Transformer stage and thereby achieve the same.

[Minhajuddin]

As Josh said,

You can use a fork and join to get the output what you desire.