r-expert

name	r-expert
version	1.0.0
description	Expert-level R statistical computing, data analysis, and visualization
category	languages
tags	["r","statistics","data-analysis","ggplot2","tidyverse"]
allowed-tools	["Read","Write","Edit","Bash(R:*, Rscript:*)"]

R Statistical Computing Expert

Expert guidance for R programming, statistical analysis, data visualization, and data science.

Core Concepts

R Fundamentals

• Vectors and data frames
• Factors and lists
• Functions and apply family
• Packages and libraries
• R Markdown
• Tidyverse ecosystem

Statistical Analysis

• Descriptive statistics
• Hypothesis testing
• Regression analysis
• ANOVA
• Time series analysis
• Machine learning

Data Visualization

• ggplot2
• Base R graphics
• Interactive plots (plotly)
• Statistical charts
• Maps and spatial data

R Basics

# Vectors
numbers <- c(1, 2, 3, 4, 5)
names <- c("Alice", "Bob", "Charlie")

# Data frames
df <- data.frame(
  id = 1:5,
  name = c("Alice", "Bob", "Charlie", "David", "Eve"),
  age = c(25, 30, 35, 28, 32),
  salary =     



dfdfage     
df    


calculate_mean  x 
  x  x



sapplydfage x x  
lapply  


 meandfage   
  print
  
  print



 i  nrowdf 
  printdfnamei

Tidyverse

library(dplyr)
library(tidyr)
library(stringr)

# dplyr operations
df %>%
  filter(age > 28) %>%
  select(name, age, salary) %>%
  mutate(
    salary_bonus = salary * 1.1,
    age_group = case_when(
      age < 30 ~ "Young",
      age < 35 ~ "Mid-career",
      TRUE ~ "Senior"
    )
  ) %>%
  arrange(desc(salary)) %>%
  group_by(age_group) %>%
  summarise(
    count = n(),
    avg_salary = meansalary
    total_salary  salary
  


wide_data  data.frame
  id  
  year_2021    
  year_2022    



long_data  wide_data 
  pivot_longer
    cols  starts_with
    names_to  
    values_to  
    names_prefix  
  


wide_again  long_data 
  pivot_wider
    names_from  year
    values_from  value
    names_prefix  
  


df 
  mutate
    name_upper  str_to_uppername
    name_length  str_lengthname
    first_letter  str_subname  
  


df1  data.frameid   value1    
df2  data.frameid   value2    

inner_joindf1 df2 by  
left_joindf1 df2 by  
full_joindf1 df2 by  

ggplot2 Visualization

library(ggplot2)

# Basic scatter plot
ggplot(df, aes(x = age, y = salary)) +
  geom_point(size = 3, color = "blue") +
  geom_smooth(method = "lm", se = TRUE) +
  labs(
    title = "Age vs Salary",
    x = "Age (years)",
    y = "Salary ($)"
  ) +
  theme_minimal()

# Bar plot with facets
ggplot(df, aes(x = name, y = salary, fill = age_group)) +
  geom_col() +
  facet_wrap age_group 
  themeaxis.text.x  element_textangle   hjust  


ggplotdf aesx  age_group y  salary 
  geom_boxplotfill   
  geom_jitterwidth   alpha  


ggplotdf aesx  salary 
  geom_histogramaesy  ..density.. bins   fill   
  geom_densitycolor   size  


ggplottime_series_df aesx  date y  value 
  geom_linecolor   
  geom_point 
  scale_x_datedate_breaks   date_labels   
  themeaxis.text.x  element_textangle   hjust  

Statistical Analysis

# Descriptive statistics
summary(df)
mean(df$age)
median(df$salary)
sd(df$age)
var(df$salary)
quantile(df$age, probs = c(0.25, 0.5, 0.75))

# Correlation
cor(df$age, df$salary)
cor.test(df$age, df$salary)

# T-test
t.test(df$salary ~ df$gender)

# ANOVA
model <- aov(salary ~ age_group, data = df)
summary(model)
TukeyHSD(model)


lm_model  lmsalary  age  experience data  df
summarylm_model


new_data  data.frameage    experience   
predictlm_model new_data interval  


multi_model  lmsalary  age  experience  education data  df
summarymulti_model


parmfrow   
plotmulti_model


logit_model  glmoutcome  age  salary
                   data  df
                   family  binomiallink  
summarylogit_model

Time Series Analysis

library(forecast)

# Create time series
ts_data <- ts(data, start = c(2020, 1), frequency = 12)

# Decomposition
decomposed <- decompose(ts_data)
plot(decomposed)

# ARIMA model
auto_arima <- auto.arima(ts_data)
summary(auto_arima)

# Forecasting
forecast_result <- forecast(auto_arima, h = 12)
plot(forecast_result)

# Accuracy metrics
accuracy(forecast_result)

Machine Learning

library(caret)
library(randomForest)

# Split data
set.seed(123)
train_index <- createDataPartition(df$outcome, p = 0.8, list = FALSE)
train_data <- df[train_index, ]
test_data <- df[-train_index, ]

# Train model
rf_model <- randomForest(
  outcome ~ .,
  data = train_data,
  ntree = 500,
  importance = TRUE
)

# Predictions
predictions <- predict(rf_model, test_data)

# Confusion matrix
confusionMatrix(predictions, test_data$outcome)

# Feature importance
importance(rf_model)
varImpPlotrf_model


train_control  trainControl
  method  
  number  
  savePredictions  


cv_model  train
  outcome  .
  data  train_data
  method  
  trControl  train_control


printcv_model

R Markdown

---
title: "Analysis Report"
author: "Data Scientist"
date: "`r Sys.Date()`"
output:
  html_document:
    toc: true
    toc_float: true
    code_folding: hide
---

## Introduction

This analysis explores the relationship between variables.

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE, message = FALSE, warning = FALSE)
library(tidyverse)

Data Loading

df <- read.csv("data.csv")
head(df)

Visualization

ggplot(df, aes(x = x, y = y)) +
  geom_point() +
  theme_minimal()

Results

The analysis shows that r cor(df$x, df$y) correlation.

## Data Import/Export

```r
# CSV
df <- read.csv("data.csv")
write.csv(df, "output.csv", row.names = FALSE)

# Excel
library(readxl)
library(writexl)
df <- read_excel("data.xlsx", sheet = "Sheet1")
write_xlsx(df, "output.xlsx")

# JSON
library(jsonlite)
df <- fromJSON("data.json")
write_json(df, "output.json")

# Database
library(DBI)
library(RSQLite)
con <- dbConnect(SQLite(), "database.db")
df <- dbReadTable(con, "table_name")
dbWriteTable(con, "new_table", df)
dbDisconnect(con)

# Web APIs
library(httr)
response <- GET("https://api.example.com/data")
data <- content(response, as = "parsed")

Best Practices

Code Style

• Use <- for assignment
• Follow tidyverse style guide
• Write functions for repeated code
• Use meaningful variable names
• Comment complex operations
• Use %>% pipe for readability

Data Analysis

• Always explore data first
• Check for missing values
• Validate assumptions
• Use visualization
• Document your analysis
• Make analysis reproducible

Performance

• Vectorize operations
• Use data.table for large data
• Avoid growing objects in loops
• Profile code with Rprof()
• Use parallel processing
• Cache expensive computations

Anti-Patterns

❌ Growing vectors in loops ❌ Not setting random seed ❌ Ignoring NA values ❌ Using attach() ❌ Not documenting code ❌ Hardcoding file paths ❌ Not checking assumptions

Resources

• R Documentation: https://www.r-project.org/
• Tidyverse: https://www.tidyverse.org/
• ggplot2: https://ggplot2.tidyverse.org/
• R for Data Science (book): https://r4ds.had.co.nz/
• CRAN Task Views: https://cran.r-project.org/web/views/