Package 'coldrift'

Title: Lightweight Column Drift Detection for Tabular Data
Description: Provides simple and efficient methods to detect column-level data drift between reference and target datasets. Designed for monitoring tabular data pipelines and machine learning inputs using statistical distance measures.
Authors: Yash Auti [aut, cre]
Maintainer: Yash Auti <[email protected]>
License: MIT + file LICENSE
Version: 0.1.1
Built: 2026-06-03 07:24:44 UTC
Source: https://github.com/cran/coldrift

Help Index

Detect Column-Level Data Drift

Description

Compares reference and target datasets to identify column-level drift using statistical distance measures.

Usage

detect_column_drift(reference, target, alpha = 0.05)

Arguments

reference

A data.frame representing baseline data
target

A data.frame representing new incoming data
alpha

Significance level for drift detection

Value

A data.frame with drift statistics per column

Examples

ref <- data.frame(
  age = c(25, 30, 35, 40),
  city = c("A", "B", "A", "C")
)

new <- data.frame(
  age = c(26, 31, 36, 41),
  city = c("A", "B", "C", "C")
)

detect_column_drift(ref, new)