Causal Analysis in R: Implementing DAGs with a Simple Example

Readme.txt

This R script provides a step-by-step tutorial on implementing DAGs in R for causal inference. We use the dagitty and ggdag packages with a simple simulated dataset containing three variables: Y (outcome), A (exposure), and L (confounder).
A step by step video demonstration can be found at:  https://youtu.be/12gb1GQCd4I 

R Script

### How to Implement Directed Acyclic Graphs for Causal Inference in R:  The Simple Case

##A Directed Acyclic Graph (DAG) is a finite graph composed of nodes connected 
##     by directed edges with no cycles, meaning it's impossible to start at any node 
##     and follow a sequence of edge directions that returns to the same node.
##in DAGs, nodes to be arranged in a linear order that respects the direction of 
##     edges and represents dependencies or causal relationships.
##Directed Acyclic Graphs (DAGs) are used in causal inference to visually represent 
##     assumptions about causal relationships, identify confounders, guide variable 
##     selection for adjustment, and clarify the pathways affecting causal effect estimation

##Generate a sample data set
set.seed(09072025)
n <- 500
L <- rnorm(n)                                 # Confounder
A <- 0.8 * L + rnorm(n)                       # Exposure: influenced by L
Y <- 1.2 * A + 0.7 * L + rnorm(n)             # Outcome: influenced by A and L
dat <- data.frame(L, A, Y)
head(dat)

##Specify and plot the DAG
# Install if not already installed
#install.packages(c("dagitty", "ggdag"))

library(dagitty)
library(ggdag)

# Define the DAG: L -> A, L -> Y, A -> Y
dag <- dagitty("dag {
  L -> A
  L -> Y
  A -> Y
}")

# Plot the DAG
ggdag(dag) + theme_dag()

##Analyze DAG: Find adjustment sets
# Find minimal adjustment sets for estimating effect of A on Y
adjustmentSets(dag, exposure = "A", outcome = "Y")

##Regression for Causal Effect
# Estimate causal effect, adjusting for confounder L
model <- lm(Y ~ A + L, data = dat)
summary(model)

VideoDemo

A step by step video demonstration can be found at:  https://youtu.be/12gb1GQCd4I  

A step by step video demonstration can be found at: https://youtu.be/12gb1GQCd4I