How to Update Metric Card Visualization on User Filter List

Ask the Community
1

Hi all looking for some advice.

So let’s say I have a metric I want to display called “Mean Duration Minutes” in a metric card. Calculating the stat itself is very straightforward, take the input set, filter on user filter list, and then aggregate the mean.

But what if I wanted to display the visualization in the metric card?

Metric cards have an option to show a visualization which takes in a time series set (with one object) and then you can choose a property to visualize. The thing is I’m having trouble finding a way so that the incoming time series property is then filtered by the filter list, changing the visualization depending on the current active filters.

I guess the question boils down to: How can you use a filter list to filter incoming time series property data?

Thanks for reading and any advice is appreciated

2

Metric card visualizations need a time series set derived downstream of your filter, not before it. If it’s static despite filter changes, you’re deriving the time series too early in the variable chain.

The problem:

Metric cards take a time series set variable, not a filtered object set. If you wire a visualization directly to an unfiltered source, it won’t react when your filter list changes — the metric updates but the sparkline stays static.

The fix:

  1. Filter List outputs an object set filter variable

  2. Apply that filter to your base object set → Filtered Object Set variable

  3. From the filtered set, derive a Time Series Set variable (select your time series property, e.g., “Duration Minutes”)

  4. Feed that derived time series set into the metric card’s visualization config

Variable chain:

Base Object Set
  → Filter List (outputs filter)
    → Filtered Object Set (base + filter applied)
      → Time Series Set (derived from filtered set)
        → Metric Card visualization input

Secondary constraint is aggregation:

Metric card visualizations expect a time series set from a small number of objects (docs say “one object”). If your filtered set contains many objects, you have two paths:

  1. Select a single object — add a dropdown or selector before deriving the time series.

  2. Aggregate programmatically — build a TypeScript Function that takes the filtered object set, computes a rolled-up metric (e.g., mean duration over time), and returns a single-object time series. Wire that function output into the visualization.

The sparkline isn’t natively built for “show aggregated trend across N objects” — it’s built for “show time series for one object.”

Docs: