Data Warehousing in QlikSense

[ravindraqv9]

Hi Team,

Hope everyone is doing good.

I am creating a dashboard in QlikSense with nearly 40 Transaction files. Client providing all the transaction tables in a CSV format.

Now i want to know is there any best way to split the transaction tables into dimension tables and fact tables using QlikSense as i don't have any data mart available for this purpose?

The option which i am thinking is to concatenate all the files but the issue is at last i will at least have 1000+ columns in the table.

Please suggest me the best possible way to handle this.

Thanks,

Ravi

[Somasundaram]

Hi,

Don't concatenate all the tables. Try to understand the relationship between data/tables.

Try to reduce the tables using joins, Mapping, resident, preceding loads based on the data.

and create a data model.

[ravindraqv9]

Hi Somasundaram,

Thanks for your reply. There is no relation between the tables as they are metrics of different processes. All of them transaction tables with few (2 or 3)common dimensions across them. So i can't use joins or Mappings.

Best Regards,

Ravi

[timpoismans]

Hi Ravi

If the transaction tables are nothing alike, but you want to create a couple of dashboards, you'll have to build a sort of datamodel so that all of the data gets displayed correctly. (Otherwise, you're building 40 different versions of the same dashboard most likely)

Things to consider would be:

• Do you need all the fields (columns)
• Which fields have the same type of value, but have different names
• ...

Things like this will help you solve the issue at hand and consolidate the 40 different data sources into a workable datamodel.

Regards,

Tim P.