Packages, Functions, and Best Practices
The Comprehensive R Archive Network (CRAN) continues to grow rapidly, with thousands of packages available for various statistical and data science applications. As of 2025, CRAN hosts over 20,000 packages, making R one of the most comprehensive statistical computing environments.
Recent trends in R package development include:
The tidyverse represents a modern approach to data science in R, emphasizing:
# Install the complete tidyverse
# install.packages("tidyverse")
# Core packages included:
library(dplyr) # Data manipulation
library(ggplot2) # Data visualization
library(tidyr) # Data tidying
library(readr) # Data import
library(purrr) # Functional programming
library(tibble) # Modern data frames
library(stringr) # String manipulation
library(forcats) # Factor handling
library(nycflights13)The dplyr package provides a grammar of data manipulation with five key verbs:
# Example using starwars dataset
library(dplyr)
starwars %>%
filter(species == "Human") %>%
select(name, height, mass, homeworld) %>%
mutate(bmi = mass / (height/100)^2) %>%
arrange(desc(bmi)) %>%
slice_head(n = 5)# A tibble: 5 × 5
name height mass homeworld bmi
<chr> <int> <dbl> <chr> <dbl>
1 Owen Lars 178 120 Tatooine 37.9
2 Darth Vader 202 136 Tatooine 33.3
3 Beru Whitesun Lars 165 75 Tatooine 27.5
4 Wedge Antilles 170 77 Corellia 26.6
5 Luke Skywalker 172 77 Tatooine 26.0# Computing row-wise statistics
df <- tibble(x = 1:3, y = 3:5, z = 5:7)
df %>%
rowwise() %>%
mutate(row_mean = mean(c(x, y, z)))# A tibble: 3 × 4
# Rowwise:
x y z row_mean
<int> <int> <int> <dbl>
1 1 3 5 3
2 2 4 6 4
3 3 5 7 5# Apply functions across multiple columns
starwars %>%
summarise(across(where(is.numeric),
list(mean = ~mean(.x, na.rm = TRUE),
sd = ~sd(.x, na.rm = TRUE))))# A tibble: 1 × 6
height_mean height_sd mass_mean mass_sd birth_year_mean birth_year_sd
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 175. 34.8 97.3 169. 87.6 155.# Joining datasets
flights %>%
left_join(airlines, by = "carrier") %>%
left_join(weather, by = c("year", "month", "day", "hour", "origin"))# A tibble: 336,776 × 30
year month day dep_time sched_dep_time dep_delay arr_time sched_arr_time
<int> <int> <int> <int> <int> <dbl> <int> <int>
1 2013 1 1 517 515 2 830 819
2 2013 1 1 533 529 4 850 830
3 2013 1 1 542 540 2 923 850
4 2013 1 1 544 545 -1 1004 1022
5 2013 1 1 554 600 -6 812 837
6 2013 1 1 554 558 -4 740 728
7 2013 1 1 555 600 -5 913 854
8 2013 1 1 557 600 -3 709 723
9 2013 1 1 557 600 -3 838 846
10 2013 1 1 558 600 -2 753 745
# ℹ 336,766 more rows
# ℹ 22 more variables: arr_delay <dbl>, carrier <chr>, flight <int>,
# tailnum <chr>, origin <chr>, dest <chr>, air_time <dbl>, distance <dbl>,
# hour <dbl>, minute <dbl>, time_hour.x <dttm>, name <chr>, temp <dbl>,
# dewp <dbl>, humid <dbl>, wind_dir <dbl>, wind_speed <dbl>, wind_gust <dbl>,
# precip <dbl>, pressure <dbl>, visib <dbl>, time_hour.y <dttm>R TaskView is a curated collection of packages organized by subject area, maintained by domain experts. These views help users navigate the vast ecosystem of R packages by providing structured recommendations for specific analytical domains.
# Common data import packages
library(readr) # CSV, TSV files
library(readxl) # Excel files
library(haven) # SPSS, Stata, SAS
library(DBI) # Databases
library(jsonlite) # JSON data
library(xml2) # XML data# Visualization ecosystem
library(ggplot2) # Grammar of graphics
library(plotly) # Interactive plots
library(leaflet) # Interactive maps
library(DT) # Interactive tables
library(htmlwidgets) # Web-based widgetsA function in R is a set of statements organized together to perform a specific task. Functions are fundamental building blocks that allow you to:
function_name <- function(argument1, argument2 = default_value) {
# Function body
result <- some_computation(argument1, argument2)
return(result) # Optional - R returns last expression
}# Example function with different argument types
calculate_stats <- function(x, na.rm = FALSE, ...) {
if (na.rm) {
x <- x[!is.na(x)]
}
list(
mean = mean(x, ...),
median = median(x, ...),
sd = sd(x, ...)
)
}
# Usage
data <- c(1, 2, 3, NA, 5)
calculate_stats(data, na.rm = TRUE)$mean
[1] 2.75
$median
[1] 2.5
$sd
[1] 1.707825R matches arguments in three ways:
# All equivalent calls
mean(x = c(1, 2, 3), na.rm = TRUE)[1] 2mean(c(1, 2, 3), na.rm = TRUE)[1] 2mean(c(1, 2, 3), na = TRUE) # Partial matching[1] 2# Get help for a function
?mean
help(mean)
# Search for functions
??regression # Fuzzy search
help.search("regression")
# View package documentation
help(package = "dplyr")
# View vignettes
vignette("dplyr")
browseVignettes("dplyr")# Good: Function does one thing well
calculate_bmi <- function(weight_kg, height_m) {
bmi <- weight_kg / (height_m^2)
return(bmi)
}
# Bad: Function tries to do too many things
calculate_everything <- function(weight_kg, height_m) {
bmi <- weight_kg / (height_m^2)
category <- if (bmi < 18.5) "Underweight" else "Normal"
plot(weight_kg, height_m) # Side effect!
return(list(bmi = bmi, category = category))
}# Good
calculate_confidence_interval <- function(data, confidence_level = 0.95) {
# Function implementation
}
# Bad
ci_calc <- function(d, cl = 0.95) {
# Function implementation
}safe_divide <- function(x, y) {
# Input validation
if (!is.numeric(x) || !is.numeric(y)) {
stop("Both x and y must be numeric")
}
if (y == 0) {
warning("Division by zero, returning Inf")
return(Inf)
}
return(x / y)
}#' Calculate Body Mass Index
#'
#' This function calculates BMI from weight and height measurements
#'
#' @param weight_kg Numeric vector of weights in kilograms
#' @param height_m Numeric vector of heights in meters
#' @return Numeric vector of BMI values
#' @examples
#' calculate_bmi(70, 1.75)
#' calculate_bmi(c(70, 80), c(1.75, 1.80))
#' @export
calculate_bmi <- function(weight_kg, height_m) {
if (length(weight_kg) != length(height_m)) {
stop("weight_kg and height_m must have the same length")
}
bmi <- weight_kg / (height_m^2)
return(bmi)
}# Header with script information
# Title: Data Analysis Pipeline
# Author: Your Name
# Date: 2025-01-01
# Purpose: Analyze survey data and generate report
# Load required packages
library(tidyverse)
library(here)
library(scales)
# Set global options
options(stringsAsFactors = FALSE)
# Define constants
SIGNIFICANCE_LEVEL <- 0.05
OUTPUT_DIR <- here("output")
# Source custom functions
# source(here("R", "helper_functions.R"))
# Main analysis code...# Robust data reading
read_data_safely <- function(file_path) {
tryCatch({
data <- read_csv(file_path)
message(paste("Successfully loaded", nrow(data), "rows"))
return(data)
}, error = function(e) {
stop(paste("Failed to read file:", file_path, "\nError:", e$message))
})
}data <- read_csv('data.csv')# Bad: Repetitive code
data$age[data$age < 0] <- NA
data$age[data$age > 120] <- NA
data$income[data$income < 0] <- NA
data$income[data$income > 1000000] <- NA
data$height[data$height < 50] <- NA
data$height[data$height > 250] <- NA# Good: Reusable function
clean_outliers <- function(x, min_val, max_val) {
x[x < min_val | x > max_val] <- NA
return(x)
}
# Usage
data <- data %>%
mutate(
age = clean_outliers(age, 0, 120),
income = clean_outliers(income, 0, 1000000),
height = clean_outliers(height, 50, 250)
)
# Even better: Use across() for multiple columns
outlier_rules <- list(
age = c(0, 120),
income = c(0, 1000000),
height = c(50, 250)
)
clean_multiple_outliers <- function(data, rules) {
for (col in names(rules)) {
if (col %in% names(data)) {
data[[col]] <- clean_outliers(data[[col]], rules[[col]][1], rules[[col]][2])
}
}
return(data)
}
data_clean <- clean_multiple_outliers(data, outlier_rules)# Bad: Repetitive plotting code
ggplot(data, aes(x = age)) +
geom_histogram(bins = 30) +
theme_minimal() +
labs(title = "Distribution of Age")
ggplot(data, aes(x = income)) +
geom_histogram(bins = 30) +
theme_minimal() +
labs(title = "Distribution of Income")
ggplot(data, aes(x = height)) +
geom_histogram(bins = 30) +
theme_minimal() +
labs(title = "Distribution of Height")
# Good: Reusable plotting function
create_histogram <- function(data, variable, bins = 30) {
ggplot(data, aes(x = .data[[variable]])) +
geom_histogram(bins = bins, fill = "steelblue", alpha = 0.7) +
theme_minimal() +
labs(
title = paste("Distribution of", str_to_title(variable)),
x = str_to_title(variable),
y = "Count"
)
}
# Usage
variables <- c("age", "income", "height")
plots <- map(variables, ~create_histogram(data, .x))
names(plots) <- variables
# Even better: Create all plots at once
create_histograms_for_numeric <- function(data) {
numeric_vars <- select(data, where(is.numeric)) %>% names()
map(numeric_vars, ~create_histogram(data, .x)) %>%
set_names(numeric_vars)
}
all_plots <- create_histograms_for_numeric(data)# Bad: Repetitive summary calculations
age_summary <- data %>%
summarise(
mean = mean(age, na.rm = TRUE),
median = median(age, na.rm = TRUE),
sd = sd(age, na.rm = TRUE),
min = min(age, na.rm = TRUE),
max = max(age, na.rm = TRUE)
)
income_summary <- data %>%
summarise(
mean = mean(income, na.rm = TRUE),
median = median(income, na.rm = TRUE),
sd = sd(income, na.rm = TRUE),
min = min(income, na.rm = TRUE),
max = max(income, na.rm = TRUE)
)# Good: Using across() to avoid repetition
summary_stats <- data %>%
summarise(across(
where(is.numeric),
list(
mean = ~mean(.x, na.rm = TRUE),
median = ~median(.x, na.rm = TRUE),
sd = ~sd(.x, na.rm = TRUE),
min = ~min(.x, na.rm = TRUE),
max = ~max(.x, na.rm = TRUE)
),
.names = "{.col}_{.fn}"
))
# Or create a reusable function
get_summary_stats <- function(data, variables = NULL) {
if (is.null(variables)) {
variables <- select(data, where(is.numeric)) %>% names()
}
data %>%
select(all_of(variables)) %>%
summarise(across(
everything(),
list(
n = ~sum(!is.na(.x)),
mean = ~mean(.x, na.rm = TRUE),
median = ~median(.x, na.rm = TRUE),
sd = ~sd(.x, na.rm = TRUE),
min = ~min(.x, na.rm = TRUE),
max = ~max(.x, na.rm = TRUE)
),
.names = "{.col}_{.fn}"
)) %>%
pivot_longer(everything(), names_to = "statistic", values_to = "value") %>%
separate(statistic, into = c("variable", "stat"), sep = "_(?=[^_]*$)")
}
# Usage
summary_table <- get_summary_stats(data, c("age", "income", "height"))Following these principles will help you write more maintainable, readable, and efficient R code. Remember:
The combination of good function design, proper script organization, and DRY principles will make your R code more professional and easier to maintain over time.