Choose a chart with AI
Turn an analytical question and a set of columns into a chart specification with a reason for each visual choice.
Choose a chart with AI when a table supports several different stories. A list of delivery times might call for a distribution, but a question about whether larger orders take longer calls for a relationship between two measures. The chart-chooser skill asks for the comparison before assigning an axis.
The output is a design specification. It can say which column belongs on each axis, how marks should be ordered and what units readers need. A second candidate and its trade-off give the person building the chart a practical alternative when the layout or audience changes.
What it does
- Distinguish a ranking question from a part-to-whole question.
- Specify a layout for comparing two time points across regions.
- Choose a distribution view for measurements with a wide spread.
- Check the baseline, units, ordering and labels before handing work to a designer.
- Compare two plausible chart types and explain the trade-off to a colleague.
Install and use
- Give your agent the question and column descriptions. The optional local CSV helper requires Python 3 and reads the file you choose.
- Explain the audience and the intended comparison. Identify units, dates or categories with an order of their own.
- Use the resulting specification in your preferred charting tool and review the labels at their final size.
FAQ
Source
Made by Jannik Reinhard. Original skill: chart-chooser, published on GitHub under the MIT licence.
View on GitHub