Network Causal Tree

The NetworkCausalTree package introduces a machine learning method that uses tree-based algorithms and an Horvitz-Thompson estimator to assess the heterogeneity of treatment and spillover effects in clustered network interference. Causal inference studies typically assume no interference between individuals, but in real-world scenarios where individuals are interconnected through social, physical, or virtual ties, the effect of a treatment can spill over to other connected individuals in the network. To avoid biased estimates of treatment effects, interference should be considered. Understanding the heterogeneity of treatment and spillover effects can help policy-makers scale up interventions, target strategies more effectively, and generalize treatment spillover effects to other populations.

Getting Started

Installing the latest developing version:

library(devtools)
install_github("fbargaglistoffi/NetworkCausalTree", ref="master")

Import:

library("NetworkCausalTree")

Examples

Example 1

Data generated using Erdos Renyi networks.

## Examples
dataset <- data_generator(N = 4000, 
                          M = 4,
                          k = 80, 
                          p = rep(0.2,4000), 
                          het = TRUE, 
                          h = 2, 
                          method_networks = "er", 
                          param_er = 0.1)

Singular splitting based on the main treatment effect only

result <- NetworkCausalTree(X = dataset[["X"]],
                            Y = dataset[["Y"]],
                            W = dataset[["W"]], 
                            A = dataset[["A"]],
                            K = dataset[["K"]],
                            p = dataset[["p"]], 
                            effect_weights = c(1,0,0,0),
                            ratio_disc = 0.5,
                            depth = 3,
                            minsize = 5, 
                            method = "singular",
                            output = "estimation")


title <- expression(paste("CAUSAL TREE TARGETED TO ",tau,"(1,0;0,0)"),sep="")
cov_names <- colnames(dataset[["X"]])

plot_NCT(NCT = result, 
         cov_names = cov_names,
         title = title)

Example 2

Data generated using Barabasi - Albert networks.

dataset <- data_generator(N = 4000,
                          M = 4,
                          k = 80,
                          p = rep(0.2,4000),
                          het = TRUE,
                          h = 3,
                          method_networks = "sf")

Composite splitting (NCT based on all the four effects)

result <- NetworkCausalTree(X = dataset[["X"]],
                            Y = dataset[["Y"]],
                            W = dataset[["W"]],
                            A = dataset[["A"]],
                            K = dataset[["K"]],
                            p = dataset[["p"]],
                            effect_weights =   c(0.25, 0.25, 0.25, 0.25),
                            ratio_disc = 0.5,
                            depth = 2,
                            minsize = 5,
                            method = "composite",
                            output = "detection")

title <- expression("CAUSAL TREE TARGETED TO ALL THE EFFECTS")
cov_names <- colnames(dataset[["X"]])

plot_NCT(NCT = result, 
         cov_names = cov_names,
         title = title,
         output = "detection")

Code of Conduct

Please note that the CRE project is released with a Contributor Code of Conduct. By contributing to this project, you agree to abide by its terms. More information about the opening issues and contributing (i.e., git branching model) can be found here.

Cite

@article{bargagli2024heterogeneous,
  title={Heterogeneous Treatment and Spillover Effects under Clustered Network Interference},
  author={Bargagli-Stoffi, Falco J and Tortu, Costanza and Forastiere, Laura},
  journal={Annals of Applied Statistics},
  year={2024}
}