cleanR: Streamlined Data Cleaning in R

Overview

cleanR is an R package designed to simplify and streamline the data cleaning process. It provides a unified interface for common data preparation tasks including:

• Handling missing values (removal or imputation)
• Removing duplicate rows
• Standardizing categorical variables
• Converting column data types
• Detecting and removing outliers
• Reshaping data between long and wide formats
• Applying mathematical column transformations
• Generating cleaning reports to see exactly what changed

Installation

You can install the development version of cleanR from GitHub with:

# install.packages("devtools")
devtools::install_github("Steven-Nanga/cleanR")

Features

Function	Description
handle_missing()	Handle missing values (remove, impute with mean/median/mode, or fill with constant)
remove_duplicates()	Remove duplicate rows from a data frame
standardize_categories()	Trim whitespace, lowercase, and normalize categorical variables
convert_types()	Convert column types (numeric, integer, character, factor, date, logical)
remove_outliers()	Remove outliers using IQR or z-score methods
reshape_data()	Convert between long and wide data formats
transform_cols()	Apply transformations: log, sqrt, square, normalize, standardize, and more
clean_data()	A wrapper that applies multiple cleaning steps in one call (with optional report)
compare_data()	Compare two data frames and generate a detailed report of differences

Quick Start

library(cleanR)

# Sample data with common issues
data <- data.frame(
  id = c(1, 2, 2, 3, 4, 5),
  category = c("Cat ", "DOG", "dog", "FISH", "Bird", "CAT"),
  value = c(10, NA, 20, 100, 40, 50),
  date = c("2023-01-01", "2023-02-01", "2023-02-01", "2023-03-01", "2023-04-01", "2023-05-01"),
  stringsAsFactors = FALSE
)

# Clean everything in one call
cleaned_data <- clean_data(
  data,
  missing_cols = "value",
  missing_method = "mean",
  duplicate_cols = c("id", "category"),
  categorical_cols = "category",
  type_list = list(value = "numeric", date = "date"),
  outlier_cols = "value"
)

print(cleaned_data)

Detailed Function Usage

Handle Missing Values

handle_missing(data, cols = NULL, method = "remove", fill_value = NULL)

• data: A data frame
• cols: Columns to handle (default is all columns)
• method: One of "remove", "mean", "median", "mode", or "constant"
• fill_value: Value to use when method is "constant"

Examples:

# Remove rows with NA
handle_missing(data, method = "remove")

# Impute with column mean
handle_missing(data, cols = "value", method = "mean")

# Impute with most frequent value
handle_missing(data, cols = "category", method = "mode")

# Fill with a constant
handle_missing(data, cols = "value", method = "constant", fill_value = 0)

Remove Duplicates

remove_duplicates(data, cols = NULL)

• data: A data frame
• cols: Columns to consider when identifying duplicates (default is all columns)

Standardize Categories

standardize_categories(data, cols)

• data: A data frame
• cols: Categorical columns to standardize (trims whitespace, lowercases, replaces non-alphanumeric characters with underscores)

Convert Types

convert_types(data, type_list, date_format = NULL)

• data: A data frame
• type_list: A named list specifying column names and their desired types. Supported types: "numeric", "integer", "character", "factor", "date", "logical"
• date_format: Optional date format string (e.g. "%Y-%m-%d", "%m/%d/%Y")

Remove Outliers

remove_outliers(data, cols, method = "iqr", threshold = 1.5)

• data: A data frame
• cols: Numeric columns to check for outliers
• method: "iqr" (interquartile range) or "zscore"
• threshold: Threshold for outlier detection (default 1.5 for IQR)

Rows with NA values in the checked columns are preserved.

Reshape Data

reshape_data(data, direction, cols = NULL, names_to = "name", values_to = "value",
             names_from = NULL, values_from = NULL)

• data: A data frame
• direction: "long" or "wide"
• cols: (long) Columns to pivot into longer format
• names_to / values_to: (long) Names for the new key/value columns
• names_from / values_from: (wide) Columns to spread into wider format

Examples:

# Wide to long
wide_df <- data.frame(id = 1:3, height = c(5.5, 6.0, 5.8), weight = c(150, 180, 165))
reshape_data(wide_df, direction = "long", cols = c("height", "weight"))

# Long to wide
long_df <- data.frame(
  id = c(1, 1, 2, 2),
  measurement = c("height", "weight", "height", "weight"),
  value = c(5.5, 150, 6.0, 180)
)
reshape_data(long_df, direction = "wide", names_from = "measurement", values_from = "value")

Transform Columns

transform_cols(data, cols, method = "log", base = exp(1), new_col = FALSE)

• data: A data frame
• cols: Numeric columns to transform
• method: One of "log", "log1p", "sqrt", "square", "normalize", "standardize", or "inverse"
• base: Base for logarithm (default is natural log)
• new_col: If TRUE, keep originals and add new columns with a suffix (e.g. value_log)

Examples:

df <- data.frame(x = c(1, 10, 100, 1000))

# Natural log
transform_cols(df, "x", method = "log")

# Log base 10
transform_cols(df, "x", method = "log", base = 10)

# Min-max normalization (0 to 1)
transform_cols(df, "x", method = "normalize")

# Z-score standardization (mean 0, sd 1)
transform_cols(df, "x", method = "standardize")

# Keep original and add transformed column
transform_cols(df, "x", method = "sqrt", new_col = TRUE)

Clean Data (Wrapper Function)

clean_data(data,
           missing_cols = NULL, missing_method = "remove", missing_fill = NULL,
           duplicate_cols = NULL,
           categorical_cols = NULL,
           type_list = NULL, date_format = NULL,
           outlier_cols = NULL, outlier_method = "iqr", outlier_threshold = 1.5,
           transform_cols_list = NULL, transform_method = "log", transform_new_col = FALSE,
           report = FALSE)

Applies all cleaning steps in one call. Steps are executed in order:

1. Handle missing values
2. Remove duplicates
3. Standardize categories
4. Convert types
5. Remove outliers
6. Transform columns

Each step is skipped when its corresponding parameter is NULL.

Set report = TRUE to get a step-by-step report of every change:

result <- clean_data(
  data,
  missing_cols = "value",
  missing_method = "mean",
  duplicate_cols = c("id", "category"),
  categorical_cols = "category",
  type_list = list(date = "date"),
  outlier_cols = "value",
  report = TRUE
)

# The cleaned data
result$data

# The full report object
result$report

Compare Data (Standalone Report)

compare_data(old_data, new_data, print_report = TRUE)

• old_data: The original data frame (before cleaning)
• new_data: The modified data frame (after cleaning)
• print_report: If TRUE (default), prints the report to the console

Use this to compare any two data frames, regardless of how the cleaning was done:

original <- data
cleaned <- handle_missing(original, cols = "value", method = "median")
compare_data(original, cleaned)

The report covers:

• Dimensions: rows and columns before vs. after
• Missing values: total and per-column NA changes
• Type changes: columns whose type was converted
• Duplicates: duplicate row counts before vs. after
• Value changes: number of cells modified per column

Contributing

Contributions to cleanR are welcome! Here are some ways you can contribute:

• Report bugs and request features by opening an issue
• Submit pull requests to fix bugs or add new features
• Improve documentation or add examples
• Share your experience using cleanR

License

This project is licensed under the MIT License - see the LICENSE file for details.

Authors

• Steven Nanga

Acknowledgments

Inspired by common data cleaning challenges in R.