topic Floating-Point Precision & Type Inference between Qlik clouds in Move to Qlik Cloud Analytics

Floating-Point Precision & Type Inference between Qlik clouds

MigueldelCampo — Tue, 05 May 2026 18:41:09 GMT

When migrating to a Capacity-based tenant, users often notice that fields which previously appeared as integers now display "decimal noise" (e.g., 100 becomes 100.000000004). This isn't a bug in your data; it’s a change in how the Qlik engine interprets and stores numeric precision across different infrastructures.

Its is a real fact?

Re: Floating-Point Precision & Type Inference between Qlik clouds

rwunderlich — Wed, 06 May 2026 16:03:10 GMT

To my understanding all tenants use the same codebase and should not exhibit different behavior for identical apps. When you say "migrating to a capacity-based tenant", do you mean the apps were moved from one tenant to another? How were the apps moved?

-Rob

Re: Floating-Point Precision & Type Inference between Qlik clouds

MigueldelCampo — Wed, 06 May 2026 17:09:47 GMT

YES, this is downloading a QVF and uploading it to the other tenant.

Re: Floating-Point Precision & Type Inference between Qlik clouds

marksouzacosta — Thu, 07 May 2026 03:39:16 GMT

I think we need more details about that and this needs to be replicable.

How are you ingesting those integer values?
Does it happen 100% of times and with different apps?
How is your field tagged in both Tenants? $numeric? $numeric $integer? Blank?

Regards,

Mark Costa

Re: Floating-Point Precision & Type Inference between Qlik clouds

rwunderlich — Thu, 07 May 2026 04:11:08 GMT

A couple more questions in addition to those posed by Mark. .

1. Are you downloading/uploading using QMC, CLI, or some other method?

2. Where in the app do you see the decimal noise? Is this a chart dimension or measure? Or just when you look at the fields in a listbox?

-Rob