builder: implement edid() -- Efficient DiD estimator (no-covariate path)

.Rbuildignore

@@ -26,3 +26,20 @@ Icon?
 ^benchmark-results$
 ^dev$
 ^\.revdeplite$
+^CSX2026\.pdf$
+^Efficient_DiD\.pdf$
+^Rplots\.pdf$
+^CODEX_AUDIT_HANDOFF\.md$
+^EDID_implemention\.md$
+^IMPLEMENTATION_PLAN\.md$
+^METHODOLOGY_REVIEW\.md$
+^audit\.md$
+^audit_cov\.md$
+^comprehension\.md$
+^implementation\.md$
+^mailbox\.md$
+^spec\.md$
+^test-spec\.md$
+^data-raw$
+^updates_made\.md$
+^quality_reports$

.github/workflows/test-coverage.yml

@@ -24,10 +24,47 @@ jobs:
           needs: coverage
 
       - name: Test coverage
+        env:
+          CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
         run: |
-          covr::codecov(
-            quiet = FALSE,
-            clean = FALSE,
-            install_path = file.path(normalizePath(Sys.getenv("RUNNER_TEMP"), winslash = "/"), "package")
+          fail_root <- file.path(normalizePath(Sys.getenv("RUNNER_TEMP"), winslash = "/"), "package")
+
+          print_failures <- function() {
+            fail_files <- if (dir.exists(fail_root)) {
+              list.files(fail_root, pattern = "\\.Rout\\.fail$", recursive = TRUE, full.names = TRUE)
+            } else {
+              character()
+            }
+            for (fail_file in fail_files) {
+              cat("\n===== ", fail_file, " =====\n", sep = "")
+              cat(readLines(fail_file, warn = FALSE), sep = "\n")
+              cat("\n===== end ", fail_file, " =====\n", sep = "")
+            }
+          }
+
+          coverage <- tryCatch(
+            covr::package_coverage(
+              quiet = FALSE,
+              clean = FALSE,
+              install_path = fail_root
+            ),
+            error = function(e) {
+              print_failures()
+              stop(e)
+            }
           )
+
+          print(coverage)
+
+          codecov_token <- Sys.getenv("CODECOV_TOKEN", "")
+          if (nzchar(codecov_token)) {
+            tryCatch(
+              covr::codecov(coverage = coverage, quiet = FALSE, token = codecov_token),
+              error = function(e) {
+                warning("Codecov upload failed: ", conditionMessage(e), call. = FALSE)
+              }
+            )
+          } else {
+            message("CODECOV_TOKEN is not set; skipping Codecov upload.")
+          }
         shell: Rscript {0}

.gitignore

@@ -25,3 +25,19 @@ dev/
 .revdeplite/*.md
 .revdeplite/github/
 .revdeplite/results/
+
+# --- edid working-file ignores ---
+CSX2026.pdf
+Efficient_DiD.pdf
+Rplots.pdf
+CODEX_AUDIT_HANDOFF.md
+EDID_implemention.md
+IMPLEMENTATION_PLAN.md
+METHODOLOGY_REVIEW.md
+audit.md
+audit_cov.md
+comprehension.md
+implementation.md
+mailbox.md
+spec.md
+test-spec.md

DESCRIPTION

@@ -1,12 +1,16 @@
 Package: did
 Title: Treatment Effects with Multiple Periods and Groups
-Version: 2.5.0
+Version: 2.5.1
 Authors@R: c(person("Brantly", "Callaway", email = "brantly.callaway@uga.edu", role = c("aut", "cre")), person("Pedro H. C.", "Sant'Anna", email="pedro.santanna@emory.edu", role = c("aut")))
 URL: https://bcallaway11.github.io/did/, https://github.com/bcallaway11/did/
 Description: The standard Difference-in-Differences (DID) setup involves two periods and two groups -- a treated group and untreated group.  Many applications of DID methods involve more than two periods and have individuals that are treated at different points in time.  This package contains tools for computing average treatment effect parameters in Difference in Differences setups with more than two periods and with variation in treatment timing using the methods developed in Callaway and Sant'Anna (2021) <doi:10.1016/j.jeconom.2020.12.001>.  The main parameters are group-time average treatment effects which are the average treatment effect for a particular group at a particular time.  These can be aggregated into a fewer number of treatment effect parameters, and the package deals with the cases where there is selective treatment timing, dynamic treatment effects, calendar time effects, or combinations of these.  There are also functions for testing the Difference in Differences assumption, and plotting group-time average treatment effects.
 Depends:
     R (>= 4.1.0)
 License: GPL-3
+Copyright: The package authors, except the lookup tables in
+    inst/extdata/aks_lookup/, which are Copyright (c) 2023 Sophie Sun (MIT
+    License), vendored from the MissAdapt replication package of Armstrong,
+    Kline & Sun (2025, Econometrica). See inst/COPYRIGHTS.
 Encoding: UTF-8
 LazyData: true
 Imports:
@@ -17,6 +21,8 @@ Imports:
     DRDID (>= 1.3.0),
     generics,
     methods,
+    parallel,
+    splines,
     tidyr,
     fastglm,
     data.table (>= 1.15.4),
@@ -33,6 +39,7 @@ Suggests:
     broom,
     testthat (>= 3.0.0),
     withr,
+    callr,
     remotes,
-    callr
+    R.matlab
 Config/roxygen2/version: 8.0.0

NAMESPACE

@@ -1,5 +1,7 @@
 # Generated by roxygen2: do not edit by hand
 
+S3method(as.data.frame,edid_fit)
+S3method(coef,edid_fit)
 S3method(ggdid,AGGTEobj)
 S3method(ggdid,MP)
 S3method(glance,AGGTEobj)
@@ -8,22 +10,45 @@ S3method(nobs,AGGTEobj)
 S3method(nobs,MP)
 S3method(print,AGGTEobj)
 S3method(print,MP)
+S3method(print,edid_adaptive)
+S3method(print,edid_fit)
+S3method(print,edid_frontier)
+S3method(print,edid_hausman)
+S3method(print,edid_overid)
+S3method(print,edid_perturbation_bootstrap)
+S3method(print,edid_refit_bootstrap)
+S3method(print,edid_sargan)
 S3method(summary,AGGTEobj)
 S3method(summary,MP)
 S3method(summary,MP.TEST)
+S3method(summary,edid_fit)
 S3method(tidy,AGGTEobj)
 S3method(tidy,MP)
+S3method(vcov,edid_fit)
 export(AGGTEobj)
 export(DIDparams)
 export(MP)
 export(MP.TEST)
 export(aggte)
+export(aggte_edid)
+export(as_MP_edid)
 export(att_gt)
 export(build_sim_dataset)
 export(compute.aggte)
 export(compute.att_gt)
 export(compute.att_gt2)
 export(conditional_did_pretest)
+export(edid)
+export(edid_adaptive)
+export(edid_clear_plugin_cache)
+export(edid_frontier)
+export(edid_hausman)
+export(edid_overid)
+export(edid_perturbation_bootstrap)
+export(edid_refit_bootstrap)
+export(edid_sargan)
+export(edid_weight_plot)
+export(edid_weights)
 export(ggdid)
 export(glance)
 export(gplot)
@@ -59,6 +84,7 @@ importFrom(generics,tidy)
 importFrom(methods,as)
 importFrom(methods,is)
 importFrom(stats,aggregate)
+importFrom(stats,as.formula)
 importFrom(stats,binomial)
 importFrom(stats,complete.cases)
 importFrom(stats,cov)
@@ -74,6 +100,7 @@ importFrom(stats,predict)
 importFrom(stats,qnorm)
 importFrom(stats,quantile)
 importFrom(stats,rnorm)
+importFrom(stats,sd)
 importFrom(stats,setNames)
 importFrom(stats,var)
 importFrom(tidyr,gather)

R/AGGTEobj.R

@@ -106,11 +106,11 @@ summary.AGGTEobj <- function(object, ...) {
 
     cat("\n")
     cband_text1a <- paste0(100*(1-object$DIDparams$alp),"% ")
-    cband_text1b <- if (object$DIDparams$bstrap) {
-      if (object$DIDparams$cband) "Simult. " else "Pointwise "
-    } else {
-      "Pointwise "
-    }
+    # DIDparams$cband stores the effective band type: did's bootstrap path coerces it to FALSE
+    # whenever the simultaneous crit falls back to pointwise, and edid's analytic sup-t path sets
+    # it to TRUE when it installs a simultaneous crit without the bootstrap. Label from it directly
+    # (the old bstrap && cband rule mislabeled edid's analytic simultaneous bands as pointwise).
+    cband_text1b <- ifelse(isTRUE(object$DIDparams$cband), "Simult. ", "Pointwise ")
     cband_text1 <- paste0("[", cband_text1a, cband_text1b)
 
     cband_lower <- object$att.egt - object$crit.val.egt*object$se.egt

R/att_gt.R

@@ -302,10 +302,34 @@ att_gt <- function(yname,
   # Capture extra arguments for custom est_method
   extra_args <- list(...)
 
+  if (missing(control_group)) control_group <- "nevertreated"
+  validate_choice_scalar(
+    control_group,
+    "control_group",
+    c("nevertreated", "notyettreated"),
+    "control_group must be either 'nevertreated' or 'notyettreated'"
+  )
+  validate_choice_scalar(
+    base_period,
+    "base_period",
+    c("universal", "varying"),
+    "base_period must be either 'universal' or 'varying'."
+  )
+
   # Validate compute_inffunc (point-estimates-only switch)
   if (!is.logical(compute_inffunc) || length(compute_inffunc) != 1 || is.na(compute_inffunc)) {
     stop("compute_inffunc must be a single logical (TRUE or FALSE).")
   }
+  validate_logical_scalar(panel, "panel")
+  validate_logical_scalar(allow_unbalanced_panel, "allow_unbalanced_panel")
+  validate_logical_scalar(bstrap, "bstrap")
+  validate_logical_scalar(cband, "cband")
+  validate_logical_scalar(faster_mode, "faster_mode")
+  validate_logical_scalar(print_details, "print_details")
+  validate_logical_scalar(pl, "pl")
+  validate_positive_whole_number(cores, "cores")
+  validate_anticipation(anticipation)
+  validate_alp(alp)
   # When influence functions are not computed there are no standard errors, no
   # uniform bands, and no parallel-trends pre-test, so the bootstrap is moot.
   if (!compute_inffunc) {
@@ -356,17 +380,9 @@ att_gt <- function(yname,
     stop("Must provide idname when panel = TRUE. Set panel = FALSE for repeated cross sections.")
   }
 
-  # Validate alp (significance level)
-  if (!is.numeric(alp) || length(alp) != 1 || is.na(alp) || alp <= 0 || alp >= 1) {
-    stop("alp must be a single number strictly between 0 and 1.")
-  }
-
   # Validate biters (number of bootstrap iterations) when the bootstrap is used
   if (bstrap) {
-    if (!is.numeric(biters) || length(biters) != 1 || is.na(biters) ||
-        biters < 1 || biters != round(biters)) {
-      stop("biters must be a single positive whole number.")
-    }
+    validate_positive_whole_number(biters, "biters")
   }
 
   # Warn users about anticipation and never-treated units