Hi, How do I create a Date filter that is dynamic - for example, to be 2026 cumulative up until last complete Sunday 2 weeks ago - this would be 1st Jan until 15th Feb. it would continue to move one week forward each week (but staying fixed on the 1st Jan as a start date)

This is an interesting use case, but it doesn't map cleanly to GoodData's built-in date filter capabilities. Let me explain what's possible and where the limitations lie. What GoodData Supports GoodData offers two main dynamic date filter types: 1. Relative date filters – ranges relative to today (e.g.,

from: -6, to: 0

with

GDC.time.date

granularity = last 7 days). [Filter Visual Components] 2. To-Date filters – combine a relative period (e.g., "this year") with a

boundedFilter

that caps the end at today. For example, Year-to-Date (YTD) covers Jan 1 of the current year through today. [Using To Date filters] Why Your Use Case Is Challenging Your requirement is: • Fixed start: January 1, 2026 (a specific calendar year start) • Dynamic end: the last complete Sunday, 2 weeks ago — rolling forward weekly This combination is not directly supported out of the box. Here's why: • A Year-to-Date (YTD) filter would give you Jan 1 → today, but the end boundary would be today, not "last complete Sunday 2 weeks ago." • The

boundedFilter

in GoodData only supports bounding to

to: 0

(current period) or

from: 0

, not an offset like "2 weeks ago Sunday." [Using To Date filters] • A relative date filter with

granularity: WEEK

and

from: 0, to: -2

would give you a rolling 2-week-ago window, but the start would not be pinned to Jan 1. Closest Possible Workaround The closest native approximation would be a relative date filter via the API with a very large negative

from

value to approximate "start of year," but this would not precisely pin to Jan 1, 2026. As noted in community feedback, the UI limits the range (e.g., max 104 weeks in the UI), but the API allows larger values like

-200

. [Filter Visual Components] For a truly dynamic "Jan 1 → last complete Sunday 2 weeks ago" filter, the recommended approach would be to prepare this logic at the data source level — for example, by computing a column or flag in your database that marks whether a record falls within the desired window. GoodData's date filters are designed for analytical aggregations and do not support this level of custom boundary logic natively. [Slack community] If this is a critical requirement, I'd suggest submitting it as product feedback to GoodData, as this type of use case has been flagged by other community members as a gap in the current filtering capabilities. 👍 0 button 👎 0 button ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups. 👀 Our Support team monitors all messages and will jump in if needed.

Hi Eoin, we have reviewed the response from our AI Assistant, and can confirm that the feedback is line with the current built-in date filter capabilities - But please feel free to follow-up if you need any further assistance

Hi Eoin, We’re just wondering if you've had a chance to review the latest update from us regarding your support request? We've provided an update some time ago but we haven't heard back from you.

Hi Shamil - thanks for following up. I followed the instructions above and got it working

Thanks for getting back to us and for letting us know - Hope you have a great weekend ahead 🙂