How do I find outliers in my data with iDBQuery?

Outliers are individual records that sit far from the rest of your data, sometimes a golden insight, sometimes a data-entry error or fraud. iDBQuery surfaces them from a plain request instead of hand-written statistical SQL.

• You'd ask iDBQuery: 'Find outlier orders in the last quarter, where the value is more than three standard deviations from the mean or above the 99th percentile, and list them.'
• It computes the distribution, identifies records beyond your chosen threshold, and returns a chart plus a ranked list of the outliers.
• Each outlier is cited to its exact source row, so you can inspect the actual record, not a summarised bucket.

The practical value is deciding what an outlier means. Ask 'Are these unusually large invoices real or duplicates?' and iDBQuery keeps the thread, cross-checking against other fields, or join to a second source in the same one live model to confirm. Because it shows its SQL, an analyst can switch from a standard-deviation rule to an interquartile-range method in one follow-up. This complements anomaly detection: outlier analysis inspects individual records, anomaly detection watches metrics over time. Use it to catch data-quality problems before they reach a report, and let the Analyst agent flag the outliers most likely to be errors, with the rows to check.