Outlier_Detection_Grubb_Dixon

URS & FS — User Requirements and Functional Specification

This document describes the controlled interface, user requirements and functional behaviour of Outlier_Detection_Grubb_Dixon. The utility is intended for automated verification of laboratory, pharmacopoeial, analytical or manufacturing QC parameters using input.csv and producing a structured output.json result.

Domain limits and critical parameters

• • Outliers may arise from pipetting errors, air bubbles, instrument glitches, or sample heterogeneity.
• • Grubbs' test is most powerful for detecting a single outlier in a normal distribution.
• • Dixon's Q test is simple to calculate and widely used for quick checks of small series (3-10 replicates).
• • The utility automatically calculates statistics (G and Q) and compares them with critical tabular values.
• • "FAIL" status means the statistical hypothesis of no outliers is rejected.
• • If an outlier is detected, raw data (chromatograms, spectra) must be checked for technical errors.
• • Retesting of the series is allowed only according to approved SOPs.
• • For multiple outliers, iterative methods or other tests (e.g., Tukey's test) are required.
• • Dual check (Grubbs + Dixon) for increased reliability.
• • Calculation of cleaned mean (excluding outlier) to assess anomaly impact.
• • Report generation for LIMS archiving.
• • Grubbs G Statistic: < Critical Value
• • Dixon Q Statistic: < Critical Value
• • Alpha: 0.05 (Standard)
• • Dixon's Q Test: For small samples (3 ≤ N ≤ 10).

URS — User Requirements Specification

input.csv contract

BatchNumber,Alpha,Val1,Val2,Val3,Val4,Val5,Val6,Val7,Val8,Val9,Val10
STAT-SERIES-001,0.05,10.1,10.2,10.0,10.3,10.1,10.2,10.1,10.0,10.2,10.1
STAT-SERIES-002,0.05,10.1,10.2,10.0,10.3,10.1,12.5,10.1,10.0,10.2,10.1
STAT-SERIES-003,0.05,5.0,5.1,5.2,5.0,5.1,5.0,5.1,5.2,5.0,5.1

FS — Functional Specification

Example output.json

{
  "utilityId": "outlier-detection-grubb-dixon",
  "utilityName": "Outlier_Detection_Grubb_Dixon",
  "overallStatus": "PASS|WARNING|FAIL",
  "sourceFile": "input.csv",
  "checks": [
    {
      "parameter": "BatchNumber",
      "value": "STAT-SERIES-001",
      "status": "PASS|WARNING|FAIL",
      "message": "Rule-based check result"
    },
    {
      "parameter": "Alpha",
      "value": "0.05",
      "status": "PASS|WARNING|FAIL",
      "message": "Rule-based check result"
    },
    {
      "parameter": "Val1",
      "value": "10.1",
      "status": "PASS|WARNING|FAIL",
      "message": "Rule-based check result"
    },
    {
      "parameter": "Val2",
      "value": "10.2",
      "status": "PASS|WARNING|FAIL",
      "message": "Rule-based check result"
    },
    {
      "parameter": "Val3",
      "value": "10.0",
      "status": "PASS|WARNING|FAIL",
      "message": "Rule-based check result"
    },
    {
      "parameter": "Val4",
      "value": "10.3",
      "status": "PASS|WARNING|FAIL",
      "message": "Rule-based check result"
    },
    {
      "parameter": "Val5",
      "value": "10.1",
      "status": "PASS|WARNING|FAIL",
      "message": "Rule-based check result"
    },
    {
      "parameter": "Val6",
      "value": "10.2",
      "status": "PASS|WARNING|FAIL",
      "message": "Rule-based check result"
    },
    {
      "parameter": "Val7",
      "value": "10.1",
      "status": "PASS|WARNING|FAIL",
      "message": "Rule-based check result"
    },
    {
      "parameter": "Val8",
      "value": "10.0",
      "status": "PASS|WARNING|FAIL",
      "message": "Rule-based check result"
    },
    {
      "parameter": "Val9",
      "value": "10.2",
      "status": "PASS|WARNING|FAIL",
      "message": "Rule-based check result"
    },
    {
      "parameter": "Val10",
      "value": "10.1",
      "status": "PASS|WARNING|FAIL",
      "message": "Rule-based check result"
    }
  ],
  "criticalFindings": [],
  "warnings": [],
  "generatedFor": "QA/QC review and LIMS integration"
}

Traceability matrix

OQ/PQ test scenarios

QA/QC and change control

• Do not rename columns without updating the validator, documentation and test set.
• Retain input.csv, output.json, executable version, documentation and checksum.
• Before production use, perform IQ/OQ/PQ or equivalent CSV/CSA verification.
• Critical limits must be verified against the approved specification, local SOPs and registration dossier.