topic Is it possible to keep optimized load while performing data transformations? in QlikView

Is it possible to keep optimized load while performing data transformations?

Wed, 15 Jan 2014 13:59:07 GMT

I understand the difference between optimized vs non optimized loads, and clearly an optimized load is preferred. However, due to business requirements, data transformations in the script (such as filtering data sets, creating new calculated fields, concatenating fields for keys...etc) are inevitable. All of these transformations (except the filtering data sets using WHERE exists) force a non optimized load.

Is it possible to (for example) create a calculated field while maintaining an optimized load?

You may say that I should perform the calculations on the database side, store already transformed data into a QVD and then optimize load into a QVW. However, to maintain flexibility, I chose to perform the transformations at the QlikView level.

Any tips and tricks are appreciated.

Thanks,

Andrew

Re: Is it possible to keep optimized load while performing data transformations?

Clever_Anjos — Wed, 15 Jan 2014 14:01:40 GMT

No, every time you make a transformation you loose the 'optimized load'.

An approach is having a 3-tier QVD layer

1) Extracts

2) Do all transformations

3) load optimized

Re: Is it possible to keep optimized load while performing data transformations?

kiranmanoharrode — Wed, 15 Jan 2014 14:08:08 GMT

Dear Andrew,

It is not possible to perform Optimized load while transformation of Data;

Optimized Load is applicable only when, data is fetching from QVD. it is very useful for data filteration with where Exist() fumction.

Regards

Kiran Rode

91 8976977897

Re: Is it possible to keep optimized load while performing data transformations?

Wed, 15 Jan 2014 14:10:04 GMT

Thank you for the response Clever.

This approach is similar to what I been using and unfortunately the answer you gave is what I expected. If I am to perform transformations, I have to lose optimized load at some point.

Thanks,

Andrew