shadcn.io is not affiliated with official shadcn/ui
Data Quality Dashboard Block for Radix UI
A data quality score for an orders table with four scored dimensions, a toggle of three weightings that re-weights the overall score, a table of seven failing rules and sample bad rows in a side cell.
Installation
Overview
Judge a dataset on what matters to you, with this data quality panel for React with Radix UI. A hairline frame with a plus mark on each corner scores an Assaywell orders table on four dimensions: Completeness, Accuracy, Consistency and Timeliness, each with a Progress bar. A ToggleGroup in the header picks the weighting (Balanced, Compliance first or Freshness first), which changes each weight, each dimension's share and the overall score. Below, seven failing rules are listed by severity with the rows they affect. Pressing a field reads its rule and three of the rows that break it in the side cell. For data catalogs and pipeline monitors.
FAQ
How do I use my own scores and rules?
Edit DIMENSIONS for the four scores in tenths of a point, PRESETS for the weights (each set adds to 100) and ISSUES for the failing rules: the field, the rule, what it says, the rows, the severity and three samples. The overall score, the shares and the row totals are worked out from them.
How is the overall score worked out?
It is the mean of the four scores weighted by the chosen preset, rounded to a tenth. Each row shows its score, its weight and its share, so the shares add up to the overall. Compliance first leans on Accuracy and Consistency, Freshness first on Timeliness.
What does choosing a rule show?
Its rule as written, a sentence of what it asks, how many rows break it, the dimension it belongs to and three of those rows with the value that fails. The side cell keeps its height, so choosing another rule moves nothing.
What stays local, and how do I wire it for real?
Scores and rows are typed and nothing is profiled. Compute the scores in your quality tool, send them as DIMENSIONS, and fill samples from a sampled query of the failing rows.