Skip to content

'ratio_disc' validation #48

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
fbargaglistoffi merged 2 commits into from
Oct 13, 2026
Merged

'ratio_disc' validation #48

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
be5a703
Update NetworkCausalTree.R
charliewang123 Oct 13, 2026
0e87218
Fix subsetting of A_est to preserve dimensions
fbargaglistoffi Oct 13, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 17 additions & 5 deletions R/NetworkCausalTree.R
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -100,9 +100,22 @@ NetworkCausalTree <- function(X, Y, W,
stop('You have to specify the Adjacency Matrix A.')
}

# Ratio of clusters assigned to discovery set needs to be within boundaries
if (ratio_disc <= 0 | ratio_disc > 1) {
stop('The ratio of clusters to be assigned to the discovery set must be above 0 and below 1')
if (ratio_disc <= 0 || ratio_disc >= 1) {
stop("ratio_disc must be strictly between 0 and 1.")
}

if (k < 2) {
stop("At least two clusters are required for honest sample splitting.")
}

nclusters_disc <- round(k * ratio_disc)

if (nclusters_disc < 1) {
stop("ratio_disc is too low; no clusters are assigned to the discovery sample. Please use a higher ratio_disc.")
}

if (nclusters_disc >= k) {
stop("ratio_disc is too high; no clusters remain for the estimation sample. Please use a lower ratio_disc.")
}

# Compute network - specific information (the degree Ne, the number of treated friends Ne_treated,
Expand All @@ -127,7 +140,6 @@ NetworkCausalTree <- function(X, Y, W,
Ne = Ne)

# Sample the clusters to be assigned to the discovery set
nclusters_disc = round(k * ratio_disc)
clusters_disc <- sample(unique(K), size = nclusters_disc, replace = FALSE)

# Generate the tree on the discovery set
Expand Down Expand Up @@ -158,7 +170,7 @@ NetworkCausalTree <- function(X, Y, W,
clusters_est <- setdiff(unique(K), clusters_disc)
est_units <- which(K %in% clusters_est)

A_est <- A[est_units, est_units]
A_est <- A[est_units, est_units, drop = FALSE]
Ne_est <- rowSums(A_est)
Ne_treated_est <- as.vector(A_est %*% W[est_units])
G_est <- rep(1, length(est_units))
Expand Down
Loading