liveBook · Manning

LARGER test

As explained in Outlier Detection in Python

The LARGER test checks whether values in one column are typically larger than values in another column and flags exceptions.

The LARGER test checks whether values in one column are typically larger than values in another, then flags rows that violate this pattern. It supports explainable outlier detection by showing both the general relationship and its exceptions. section 13.3.4

A practical workflow uses DataConsistencyChecker to inspect the pattern:

  • Inspect listed example rows, exceptions, visual explanations, and interpretable outlier scores based on flagged-test counts. section 13.3.4

Abalone measurements

For the abalone dataset, Diameter is normally smaller than Length. The test found one exception, row 1210, where Diameter exceeded Length; a visualization placed the point above the main body of observations. section 13.3.4

How Outlier Detection in Python covers it

How deep it goes
Explains the comparison, exception reporting, visualizations, and scoring, but does not derive the detection algorithm.
Examples
Uses abalone Length and Diameter measurements, including the exception at row 1210 and its explanatory visualization.
What it assumes
Assumes a tabular DataFrame and a copied check_data_consistency.py file, while handling native binary, categorical, numeric, and date data.
What you can do after
Diagnose/debug tabular data with DataConsistencyChecker by locating and interpreting rows that violate a usual column relationship.
How it connects
Connects column-order exceptions with outlier detection, correlated values, sparse two-dimensional regions, and data quality.
Cover of Outlier Detection in Python
liveBookmanning.com
Read Outlier Detection in Python on liveBook Get the book at manning.com

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Outlier Detection in Python. The section links point to the parts of the book it draws on.