rDataRecon

An R package for dataset comparison and reconciliation for SDTM/ADAM data frames and Others .

Compare two data frames BASE (reference) vs COMPARE (new version) and get a detailed, structured report of every structural and value-level difference. Designed for clinical data programming, ETL validation, and general data-quality workflows.

---

Installation

# From CRAN (once published)
install.packages("rDataRecon")

# Stable version:
remotes::install_github("ganeshbabunn/rDataRecon")

# Development version:
remotes::install_github("ganeshbabunn/rDataRecon@dev")

# From local source
install.packages("../rDataRecon_X.X.tar.gz", repos = NULL, type = "source")
X indicates the version of the release.

---

Quick start

library(rDataRecon)

base_df <- data.frame(
  id   = 1:5,
  age  = c(25L, 30L, 35L, 40L, 45L),
  name = c("Sarang", "Ganesh", "Swetha", "Jyothi", "Rudresh"),
  stringsAsFactors = FALSE
)

comp_df <- data.frame(
  id   = 1:5,
  age  = c(25L, 31L, 35L, 40L, 45L),
  name = c("Sarang", "ganesh", "Swetha", "Jyothi", "Rudresh"),
  stringsAsFactors = FALSE
)

result <- recon(base_df, comp_df, id = "id")

Extract differences as a tidy data frame:

get_diffs(result)
#   obs_id variable base_value compare_value  diff
# 1  id = 2      age         30            31     1
# 2  id = 2     name        Ganesh         ganesh  NA

---

Features

Feature	Details
Zero dependencies	Base R only no additional packages required
Key-based matching	Match rows by one or more ID columns
Positional matching	Row-by-row when no ID is specified
Cross-variable mapping	Compare differently-named columns via base_var / compare_var
Absolute tolerance	criterion = 0.01 ignore differences below threshold
Relative tolerance	criterion = 0.05, relative = TRUE proportional comparison
Focused comparison	var = c("age","sbp") compare specific variables only
Report export	out = "report.txt" save full report to file
S3 result object	Class rDataRecon_result with print(), get_diffs(), summary_df()

---

Argument reference

recon() argument	Behaviour
base	Reference dataset
compare	Dataset to validate against BASE
id	Key column(s) for row matching
var	Restrict comparison to named variables
base_var + compare_var	Map differently-named columns across datasets
criterion	Numeric equality tolerance (default 1e-8)
relative = TRUE	Apply tolerance proportionally
noequal = TRUE	Suppress equal-variable output
listall = TRUE	Print all variable names in header
brief = TRUE	Summary section only
out = "file.txt"	Write report to text file

---

Examples

Cross-variable name mapping

base2 <- data.frame(id = 1:3, revenue = c(100, 200, 300))
comp2 <- data.frame(id = 1:3, sales   = c(100, 250, 300))
recon(base2, comp2, id = "id", base_var = "revenue", compare_var = "sales")

Relative tolerance (5 %)

recon(base_df, comp_df, id = "id", criterion = 0.05, relative = TRUE)

Multi-column key

recon(base_df, comp_df, id = c("site", "visit"))

Save report to file

recon(base_df, comp_df, id = "id", out = "rDataRecon_report.txt")

Programmatic use in QC pipelines

r <- recon(base_df, comp_df, id = "id", brief = TRUE)

summary_df(r)          # one-row log entry
compare_stats(base_df, comp_df)   # side-by-side numeric stats
is_structure_equal(base_df, comp_df)  # structural check

---

Report structure

recon() prints a four-section plain-text report:

========================================================================
               rDataRecon  -  Dataset Comparison Report
========================================================================
  BASE    : base_df              5 obs  x  3 vars
  COMPARE : comp_df              5 obs  x  3 vars
  Criterion : 1e-08  (absolute)
  Run time  : 2025-01-01 12:00:00
------------------------------------------------------------------------
  SECTION 1  -  Variable Structure
  SECTION 2  -  Variable Type Mismatches
  SECTION 3  -  Value Comparison
  SECTION 4  -  Summary Statistics
========================================================================
  VERDICT :  DATASETS HAVE DIFFERENCES
========================================================================

---

Return value

recon() returns an "rDataRecon_result" list invisibly:

$summary             list of counts (obs, vars, diffs)
$vars_base_only      character vector
$vars_compare_only   character vector
$type_mismatches     data.frame (variable, base_class, compare_class)
$value_diffs         data.frame (obs_id, variable, base_value,
                                  compare_value, diff)
$verdict             character string

---

Bug Reports & Feature Requests

Found a bug or have an idea to improve rDataRecon? Please raise it via the GitHub Issues tab.

When opening a new issue, use the appropriate label:

Type	Label	When to use
Bug	bug	Something is broken or behaving unexpectedly
Feature	enhancement	A new feature or improvement you'd like to see

Steps to raise an issue

1. Go to the Issues tab
2. Click "New Issue"
3. Choose the relevant template — Bug Report or Feature Request
4. Fill in the details and apply the correct label (bug or enhancement)
5. Click "Submit new issue"

Tip: Before opening a new issue, please search existing issues to avoid duplicates.

---

License

GPL-3.0 Copyright (C) 2026 Ganesh Babu G