amvcp-charts-part-of-whole

To render a donut chart for part-to-whole data visualization safely, use a fenced chart spec with a required title and series type. The runtime enforces a pie chart guardrail that automatically remaps any pie spec to a sorted bar chart to preserve ranking accuracy.

The best way to display single KPI gauges with thresholds in a dashboard is using a fenced value gauge spec. It compares a single metric against a maximum and supports optional warn or danger escalation levels for immediate visual feedback.

You create Harvey-ball qualitative rating rows by defining a fenced chart spec with the harvey-ball series type. This renders coarse five-level judgments across multiple criteria, ideal for McKinsey-style qualitative glyph rows in reports.

Standard pie charts are banned by design guardrails to prevent misuse in proportion insight. The runtime automatically remaps any pie chart spec to a sorted bar chart, ensuring ranking accuracy and enforcing a non-empty title for readability.

Yes, you need the shared dashboard infrastructure loaded alongside the amvcp design, runtime, and chart scripts. This environment provides the required palette, fence parsing, and selection atom support for agent feedback during visualization authoring.

You should not use a donut chart for data visualization when readers need precise numeric comparison rather than proportion insight or ratio breakdowns. In those cases, the design guardrails enforce remapping the spec to a sorted bar chart instead.