Hello,
After upgrade to 4.4 I have an error on Aggregate transformation. The error is
Error: Attempt to create a column with too many unique values.
Source: action “Aggregate”, module “Main”, table “…”
The input table for this aggregate transformasion has 365 mil. rows and in 4.3.1 it’s works. In 4.4 the error appere on case with 100 mil. rows on input table.
This limtation it is a bug or it is part of thouse new feautures ?
Riigh now I didn"t have a sollution for this error. The nature of data analised it causes this huge number of rows.
Best regards,
Radu.
Hi Radu,
we’re looking into the problem. I’ll keep you posted.
Hi Radu,
Do you know have many unique values are there in input dataset?
UPDATE
By unique I mean how many groups are there in result after Aggregate (you can check with old version). Also it would help a lot, if you sent us dataset with only group and aggregate columns for investigation
I was trying to simulate the aggregation tranformation loading data into QlikView. (I don’t have access at 4.3.1 version, right now).
This is log of loading script:
“in << date_oa_90d (qvd optimized) 351.732.949 lines fetched
out << in 96.346.862 lines fetched”
“in” is data input
“out” is data after grouping.
I will try to install on development enviroment also 4.3.1 version.
Thank you for providing a sample dataset. We’ve identified the problem and fixed it. The updated version has been published on our website. You can download it here: https://easymorph.com/download.html
BTW, if you haven’t done it yet - you can turn super-compression on and your project will use noticeably less RAM.
We upgraded to build 3. All good. Thank you for your support.
