Data Quality

Data Quality

What It Checks

Check	Description
Overview	Row/column count, memory usage, type distribution
Missing values	Count and percentage per column
Cardinality	Unique values, flags near-unique (possible IDs) and binary columns
Constants	Columns with only 1 unique value
Duplicates	Exact duplicate rows
High correlations	Pairs with correlation ≥ 0.95
Type inference	Detects dates and numbers stored as strings

Python API

from scomp_link import DataQualityReport

dqr = DataQualityReport(df)
report = dqr.generate()

# Access individual sections
print(report['overview'])
print(report['missing'])
print(report['constants'])
print(report['duplicates'])
print(report['correlations'])
print(report['cardinality'])

# Generate standalone HTML report
dqr.save_html('quality_report.html')

CLI

scomp-link quality --data raw_data.csv --output quality_report.html

HTML Report Contents

The generated HTML report includes:

• Summary cards (rows, columns, memory, duplicates, constants)
• Missing values table (sorted by percentage)
• Cardinality table with flags
• Type inference table
• High correlation pairs