Analytical Patterns in R

R is highly versatile for performing a wide range of analytical tasks. Below, I have outlined some common analytical patterns including data manipulation, statistical analysis, machine learning, time series analysis, and data visualization. Each section provides a brief overview and sample code.

1. Data Manipulation

The dplyr package is essential for data manipulation tasks such as filtering, selecting, mutating, and summarizing data.

Sample Code

# Load library
library(dplyr)

# Sample dataset
data <- mtcars

# Data manipulation
modified_data <- data %>%
  filter(mpg > 20) %>%  # Filter rows
  select(mpg, cyl, hp) %>%  # Select specific columns
  mutate(hp_to_wt_ratio = hp / wt) %>%  # Add new column
  summarise(avg_mpg = mean(mpg), avg_hp = mean(hp))  # Summarize data

2. Statistical Analysis

Statistical tests such as t-tests, chi-square tests, and linear regressions are common in R.

Sample Code

# Load library
library(stats)

# t-test
t_test_results <- t.test(mtcars$mpg ~ mtcars$cyl)

# Linear regression
linear_model <- lm(mpg ~ wt + hp, data = mtcars)
summary(linear_model)

3. Machine Learning

R provides packages like caret and randomForest to perform various machine learning tasks.

Sample Code

# Load libraries
library(caret)
library(randomForest)

# Sample dataset
data(iris)

# Train-Test Split
set.seed(123)
training_indices <- createDataPartition(iris$Species, p = 0.8, list = FALSE)
train_data <- iris[training_indices, ]
test_data <- iris[-training_indices, ]

# Train a Random Forest model
model <- randomForest(Species ~ ., data = train_data)

# Model prediction
predictions <- predict(model, test_data)
confusionMatrix(predictions, test_data$Species)

4. Time Series Analysis

Using packages like forecast and tsibble, R is well-suited for time series analysis and forecasting.

Sample Code

# Load libraries
library(forecast)
library(tsibble)

# Sample data
data <- AirPassengers

# Time series decomposition
decomposed <- decompose(data)
plot(decomposed)

# ARIMA model fitting
fit <- auto.arima(data)
forecast_values <- forecast(fit, h = 12)
plot(forecast_values)

5. Data Visualization

Visualizations can be created using ggplot2, one of the most powerful and flexible visualization packages in R.

Sample Code

# Load library
library(ggplot2)

# Sample dataset
data <- mtcars

# Data visualization
ggplot(data, aes(x = wt, y = mpg)) +
  geom_point(aes(color = cyl)) +  # Scatter plot with color
  geom_smooth(method = "lm", se = FALSE, color = "red") +  # Linear regression line
  labs(title = "Scatter plot of MPG vs Weight",
       x = "Weight (1000 lbs)",
       y = "Miles per Gallon")

Conclusion

R offers robust capabilities for various analytical tasks through its extensive library ecosystem:

• dplyr for data manipulation
• stats for statistical analysis
• caret and randomForest for machine learning
• forecast for time series analysis
• ggplot2 for data visualization

For more in-depth learning, consider exploring advanced courses and resources available on the Enterprise DNA Platform.

Refer back to this document as needed, and don't hesitate to experiment with your data using these patterns and techniques.