"Convert time zone" does not work for filters with dependencies

Hello everyone,

I recently ran into an issue with my data and filters.
The data I’m working with uses UTC timestamps. For usability, I want the timestamp filter to use the Berlin timezone. Therefore, I activated the “Convert time zone” feature and set it to the Berlin timezone. I can confirm that the data displayed in the visuals changed accordingly.

The problems occurred when I started filtering the data using the filters I had created. All my filters depend on the timestamp filter so that I can only select options that are actually present within the selected time frame. However, I noticed that the options in my filters don’t match the results shown in the visuals.

After some research, I found that while the visuals are correctly converted to the Berlin timezone, the filter options are still based on the same times in UTC. This creates a mismatch between the data shown in the visuals and the available filter options.

Is this a known issue that can be resolved with different settings in the analysis, or has anyone experienced this before and knows how to fix it?

Hello @andreas.teichner

Welcome to Quicksight Community and thank you for posting your question.

Could you please try this -

  • Keep the underlying data in UTC itself
  • Change the Analysis settings to Berlin time zone (refer below snippet).

Let me know if that helped you.

thank you,
Shravya

Hello @Shravya,

Thank you for reaching out.

I’m sorry if my previous explanation was misleading. The feature you mentioned is actually the one I’m having trouble with. When I enable the “Convert time zone” feature, the data in my visuals is converted correctly, but the data in my filters is not. My underlying data is all in UTC.

Let me provide a more detailed example:

The dataset includes two entries from 05:00–05:30 AM UTC with status active, and three entries from 07:00–07:30 AM UTC with status inactive.
I have two filters: one for time and one for status. The “status” filter is dependent on the date filter using the “relevant values” setting.

When I set the date filter to 07:00–07:30 AM, I see the expected entries in the visuals, and the “status” filter correctly shows only inactive.

However, when I enable the “Convert time zone” feature and set the time zone to Berlin, the visuals correctly show the data for 05:00–05:30 AM (which corresponds to 07:00–07:30 AM Berlin time), but the “status” filter still only shows inactive.

If the time zone conversion applied to the filters as well, it should show only active, since those are the entries present in the converted time range.

I hope this helps clarify the issue I’m experiencing with the “Convert time zone” feature.

Thank you for your help,
Andreas

Hi @andreas.teichner

I tried this out with a small example as below

Data in UTC in dataset

Same data in Analysis without any filters in Berlin Timezone

Same visual now with a filter on time period. The first two rows are shown per my selection which is good; but the flight id parameter control is showing all 4 rows for the date Jun 10, though I set this control to be dependent on the scheduled time and should have shown only 2 flight ids

I think this is a bug. You should report this to AWS.

Regards,
Giri

1 Like

Hey @Giridhar.Prabhu

Thank you for confirming my issue.

I also think that this is a bug and I will report this problem to the AWS support.
I will update this question when I hear back from the support or if they have a solution.

Regards,
Andreas

Hi @andreas.teichner,
It’s been awhile since we last heard from you on this thread; were you able to successfully create a support ticket and if so, did you receive feedback on this potential bug?

If we do not hear back within the next 3 business days, I’ll close out this topic.

Thank you

Hi @andreas.teichner,
Since we have not heard back further, I’ll go ahead and close out this topic. You’ll still be able to leave the results of your support ticket if you’d like. If you have any additional questions, feel free to create a new post in the community.

Thank you

Hello @Brett

sorry for not coming back sonner. I opend a ticket for the AWS support regarding this issue and they have confirmed the bug. Furthermore they confirmed that they will implement a solution for the bug, which will take some time but should be on its way.

Regards,
Andreas

3 Likes

Hi @andreas.teichner,
Thank you for sharing the update!