v2.3.1.902: bug fixes, new broom methods, and documentation improvements by pedrohcgs · Pull Request #253 · bcallaway11/did

pedrohcgs · 2026-03-02T18:18:58Z

Summary

This PR consolidates all changes on the v2.3.1.902-bugfixes branch, addressing several open issues and improving the package's robustness, usability, and documentation.

Bug fixes

Fix faster_mode crash when clustervars = c(idname, extra_var) (closes Error in mboot when clustering at two variables #175): dreamerr was enforcing scalar length before idname was stripped from clustervars; also fixed a subtle character(0)[!is.null(...)] → NA_character_ bug in att_gt.R that spuriously triggered the Wald suppression message
Add treatment irreversibility check to non-faster_mode path (closes Nested unbalanced panel data #138): both code paths now give the same clear error when treatment is not irreversible, using an efficient O(n) anyDuplicated check
Suppress Wald pre-test with informative message when extra clustering is used (closes P-value for pre-test of parallel trends assumption = 0, dynamic aggregation of ATTs not below p-val threshold pre-0 event time #186): the Wald statistic is not valid under extra clustering, so we now skip it and explain why via message()

New broom methods

nobs() for MP and AGGTEobj (closes Implement nobs #246): returns the number of unique cross-sectional units
statistic and p.value columns in tidy() (closes Please supply p-values #243): added to tidy.MP and all four aggregation types in tidy.AGGTEobj, following broom column conventions (estimate | std.error | statistic | p.value | conf.low | conf.high). p-values are pointwise/marginal and documented as not correcting for multiple testing.

Documentation

Improved roxygen docs for tidy.MP and tidy.AGGTEobj: full @return sections documenting every column, and a @details section explaining the distinction between simultaneous (conf.low/conf.high) and pointwise (point.conf.low/point.conf.high) confidence intervals

JEL replication & diagnostics (pre-existing on branch)

Download JEL replication data from GitHub when not available locally
Improved overlap warnings and factor-safe year conversion

Test plan

devtools::test() — 0 FAIL, 1 SKIP (known upstream DRDID bug), 287 PASS
devtools::check() — 0 errors, 0 warnings
Verify tidy(att_gt(...)) returns statistic and p.value columns
Verify nobs(att_gt(...)) returns number of unique units
Verify att_gt(..., clustervars = c(idname, extra_var), faster_mode = TRUE) no longer crashes

🤖 Generated with Claude Code

When clustervars includes a variable beyond idname, the analytical variance matrix V = t(inffunc) %*% inffunc / n ignores between-cluster correlation, making the Wald pre-test anti-conservative. Instead of silently producing a misleading p-value, emit a message() explaining why the test is skipped and direct users to bootstrap CIs. The test remains valid (and is reported as before) for balanced panels, unbalanced panels (influence functions aggregated to unit level), and repeated cross-sections (CLT applies per independent observation). Closes #186.

Two bugs fixed: 1. pre_process_did2.R: dreamerr's 'NULL | match' enforces scalar length, so passing clustervars = c(idname, extra_var) failed before idname could be stripped. Fix: strip idname from clustervars before the dreamerr check in validate_args, and again in the main body (validate_args works on a local copy). The non-faster_mode path (mboot.R) already handled this. 2. att_gt.R: extra_clustervars[!is.null(extra_clustervars)] is a no-op for NULL but produces NA_character_ when applied to character(0), causing a spurious Wald suppression message. Fix: remove the redundant line since the filter already produces character(0) when no extra vars remain. Closes #175.

pre_process_did.R had this check commented out as 'too slow'. Replaced the slow sapply/split approach with unique() + anyDuplicated(), which is O(n) and matches the check already in pre_process_did2.R (faster_mode). Both paths now throw the same clear error when gname is not constant within a unit across periods. Closes #138.

Document all return columns explicitly, explain the distinction between conf.low/conf.high (simultaneous bootstrap bands) and point.conf.low/point.conf.high (pointwise qnorm intervals), and add a @details section clarifying when to use each.

- Add nobs.MP and nobs.AGGTEobj S3 methods (closes #246) - Add statistic (t-stat) and p.value columns to tidy.MP and tidy.AGGTEobj following broom conventions: estimate | std.error | statistic | p.value | conf.low | conf.high (closes #243) - p.value is pointwise/marginal and documented as not accounting for multiple testing across ATT(g,t) cells - Covers all four aggregation types: dynamic, group, calendar, simple

Copilot

Pull request overview

This PR updates the did R package to address clustering and irreversibility edge cases in att_gt(), improves the Wald pre-test behavior under extra clustering, and expands broom/stats integrations (tidy() and nobs()) with clearer documentation.

Changes:

Fix faster_mode argument validation when clustervars redundantly includes idname, and add an irreversibility check to the non-faster_mode path.
Skip the Wald pre-test when extra clustering is requested (beyond idname) and inform the user.
Add nobs() methods for MP/AGGTEobj and extend tidy() outputs with statistic and p.value, plus expanded Rd documentation.

Reviewed changes

Copilot reviewed 10 out of 10 changed files in this pull request and generated 6 comments.

Show a summary per file

File	Description
`R/pre_process_did2.R`	Strips `idname` from `clustervars` before validation (fixes `faster_mode` crash) and repeats stripping in caller scope.
`R/pre_process_did.R`	Adds an O(n) irreversibility/time-invariance check for `gname` within unit.
`R/att_gt.R`	Suppresses Wald pre-test when extra clustering is used and emits an informative message.
`R/tidy.R`	Adds `nobs()` S3 methods and adds `statistic`/`p.value` columns to `tidy()` outputs with expanded roxygen docs.
`man/tidy.MP.Rd`	Updates title/description and expands return-value documentation for `tidy.MP`.
`man/tidy.AGGTEobj.Rd`	Updates title/description and expands return-value documentation + details for `tidy.AGGTEobj`.
`man/nobs.MP.Rd`	New Rd entry for `nobs.MP`.
`man/nobs.AGGTEobj.Rd`	New Rd entry for `nobs.AGGTEobj`.
`NAMESPACE`	Registers `nobs` S3 methods and imports `stats::nobs`.
`DESCRIPTION`	Bumps package version to `2.3.1.903`.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-03-02T18:42:46Z

R/tidy.R

    estimate  = x$att,
    std.error = x$se,
+    statistic = x$att / x$se,
+    p.value   = 2 * (1 - stats::pnorm(abs(x$att / x$se))),
    conf.low  = x$att - x$c * x$se,


tidy.MP() and tidy.AGGTEobj() now compute and expose statistic/p.value, but the test suite doesn’t appear to cover these new columns. Add a small testthat case that checks the columns exist and that p.value matches the expected normal-approximation calculation (including handling of NA standard errors).

Copilot · 2026-03-02T18:42:46Z

DESCRIPTION

 Package: did
 Title: Treatment Effects with Multiple Periods and Groups
-Version: 2.3.1.902
+Version: 2.3.1.903


PR title/description reference version 2.3.1.902, but DESCRIPTION bumps the package to 2.3.1.903. Please align the stated release version (either update the PR metadata/changelog text, or keep the version at 2.3.1.902 if that’s what’s intended).

Copilot · 2026-03-02T18:42:46Z

R/tidy.R

+#' @return A data frame with one row per ATT(g,t) estimate and columns:
+#'   \item{term}{ATT(g,t) label}
+#'   \item{group}{the treatment cohort g}
+#'   \item{time}{the time period t}
+#'   \item{estimate}{the ATT(g,t) point estimate}


The roxygen @return section uses \item{...}{...} directly. In the generated Rd this ends up inside \value{} without a surrounding \describe{}, which is invalid Rd markup and can break R CMD check. Wrap the column list in \describe{ ... } (or use an itemize/markdown list) and re-run roxygen so the man pages regenerate correctly.

Copilot · 2026-03-02T18:42:46Z

R/tidy.R

+#'   \item{statistic}{t-statistic (\code{estimate / std.error})}
+#'   \item{p.value}{two-sided pointwise p-value (\code{2*(1-pnorm(|t|))}).


The docs label statistic as a “t-statistic” and show pnorm(|t|), but the implementation uses a normal approximation (pnorm(abs(...))). Consider documenting this as a z/Wald statistic and use valid R syntax in the formula (e.g., abs(statistic)), to avoid misleading users.

Suggested change

#' \item{statistic}{t-statistic (\code{estimate / std.error})}

#' \item{p.value}{two-sided pointwise p-value (\code{2*(1-pnorm(|t|))}).

#' \item{statistic}{Wald z-statistic (\code{estimate / std.error})}

#' \item{p.value}{two-sided pointwise p-value (\code{2 * (1 - pnorm(abs(statistic)))}).

Copilot · 2026-03-02T18:42:47Z

R/tidy.R

+#' @return A data frame whose columns depend on \code{type}:
+#'   \item{type}{the aggregation type: \code{"simple"}, \code{"dynamic"},
+#'     \code{"group"}, or \code{"calendar"}}
+#'   \item{term}{label for each estimate}
+#'   \item{estimate}{point estimate}
+#'   \item{std.error}{standard error}
+#'   \item{statistic}{t-statistic (\code{estimate / std.error})}
+#'   \item{p.value}{two-sided pointwise p-value (\code{2*(1-pnorm(|t|))}).
+#'     Marginal per-estimate; does \strong{not} account for multiple testing
+#'     across event times or groups.}
+#'   \item{conf.low, conf.high}{simultaneous confidence band limits.  When
+#'     \code{bstrap=TRUE} and \code{cband=TRUE} these use the bootstrap uniform
+#'     critical value (\code{crit.val.egt}); otherwise they equal the pointwise
+#'     intervals.  For \code{type="simple"} and the overall average row of
+#'     \code{type="group"}, a single scalar is returned so simultaneous and
+#'     pointwise coincide.}
+#'   \item{point.conf.low, point.conf.high}{pointwise confidence interval limits
+#'     always using \code{qnorm(1 - alp/2)}.}


Same as above for tidy.AGGTEobj: the @return block uses \item{...} entries (which should be within a \describe{} list in Rd) and describes a “t-statistic” while computing p-values via the normal CDF. Adjust the roxygen markup/wording and re-run roxygen so man/tidy.AGGTEobj.Rd is valid and accurate.

Suggested change

#' @return A data frame whose columns depend on \code{type}:

#' \item{type}{the aggregation type: \code{"simple"}, \code{"dynamic"},

#' \code{"group"}, or \code{"calendar"}}

#' \item{term}{label for each estimate}

#' \item{estimate}{point estimate}

#' \item{std.error}{standard error}

#' \item{statistic}{t-statistic (\code{estimate / std.error})}

#' \item{p.value}{two-sided pointwise p-value (\code{2*(1-pnorm(|t|))}).

#' Marginal per-estimate; does \strong{not} account for multiple testing

#' across event times or groups.}

#' \item{conf.low, conf.high}{simultaneous confidence band limits. When

#' \code{bstrap=TRUE} and \code{cband=TRUE} these use the bootstrap uniform

#' critical value (\code{crit.val.egt}); otherwise they equal the pointwise

#' intervals. For \code{type="simple"} and the overall average row of

#' \code{type="group"}, a single scalar is returned so simultaneous and

#' pointwise coincide.}

#' \item{point.conf.low, point.conf.high}{pointwise confidence interval limits

#' always using \code{qnorm(1 - alp/2)}.}

#' @return A data frame with the following columns (depending on \code{type}):

#' \describe{

#' \item{type}{the aggregation type: \code{"simple"}, \code{"dynamic"},

#' \code{"group"}, or \code{"calendar"}.}

#' \item{term}{label for each estimate.}

#' \item{estimate}{point estimate.}

#' \item{std.error}{standard error.}

#' \item{statistic}{standard normal (z) test statistic

#' (\code{estimate / std.error}).}

#' \item{p.value}{two-sided pointwise p-value based on the standard normal

#' distribution (\code{2 * (1 - pnorm(|statistic|))}). Marginal

#' per-estimate; does \strong{not} account for multiple testing across

#' event times or groups.}

#' \item{conf.low, conf.high}{simultaneous confidence band limits. When

#' \code{bstrap=TRUE} and \code{cband=TRUE} these use the bootstrap uniform

#' critical value (\code{crit.val.egt}); otherwise they equal the pointwise

#' intervals. For \code{type="simple"} and the overall average row of

#' \code{type="group"}, a single scalar is returned so simultaneous and

#' pointwise coincide.}

#' \item{point.conf.low, point.conf.high}{pointwise confidence interval limits

#' always using \code{qnorm(1 - alp/2)}.}

#' }

Copilot · 2026-03-02T18:42:47Z

R/tidy.R

+nobs.MP <- function(object, ...) {
+  object$n
+}


New nobs() S3 methods were added, but there are no testthat assertions covering them. Please add tests that nobs(att_gt(...)) and nobs(aggte(...)) return the expected unit count for both faster_mode=TRUE and FALSE, so future refactors don’t silently change behavior.

…nobs tests

Copilot

Pull request overview

Copilot reviewed 11 out of 11 changed files in this pull request and generated 2 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-03-02T19:22:17Z

R/tidy.R

+#' @return Integer. The number of unique units in the data.
+#' @export
+nobs.MP <- function(object, ...) {
+  object$n


nobs.MP() returns object$n directly, but MP$n is set in att_gt() via ifelse(faster_mode, dp$id_count, dp$n), which coerces the value to double. This makes stats::nobs() return a numeric instead of an integer (and contradicts the Rd/@return). Consider coercing to integer here (or avoiding ifelse() when computing n in att_gt()).

Suggested change

object$n

as.integer(object$n)

Copilot · 2026-03-02T19:22:18Z

tests/testthat/test-tidy.R

+  xformla = ~1,
+  data    = mpdta,
+  bstrap  = TRUE,
+  cband   = TRUE


This test computes att_gt() with bstrap = TRUE and does not set biters, so it will run the default 1000 multiplier bootstrap iterations. That’s likely to make the unit test suite very slow/flaky compared to the rest of the repository’s tests (which generally set bstrap = FALSE). Consider setting bstrap = FALSE for these tidy/nobs structure checks, or explicitly setting a small biters value (and/or skip_on_cran()) if bootstrap-based output is required.

Suggested change

cband = TRUE

cband = TRUE,

biters = 20

…trap/cband=FALSE)

Copilot

Pull request overview

Copilot reviewed 11 out of 11 changed files in this pull request and generated 7 comments.

Comments suppressed due to low confidence (1)

R/pre_process_did2.R:33

dreamerr::check_set_arg(args$clustervars, "NULL | match", ...) runs before the explicit length(args$clustervars) > 1 check below, and (per the comment) enforces scalar length. That makes the custom "only provide 1 cluster variable" error effectively unreachable for multi-variable inputs. If you want the clearer error message, check length(args$clustervars) (after stripping idname) before calling check_set_arg().

  # Flag for clustervars and weightsname
  checkvar_message <- "__ARG__ must be NULL or a character scalar that is a name of a column from the dataset."
  dreamerr::check_set_arg(args$weightsname, args$clustervars, "NULL | match", .choices = data_names, .message = checkvar_message, .up = 1)

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-03-02T19:36:38Z

R/tidy.R

      std.error = c(x$overall.se, x$se.egt),
+      statistic = c(x$overall.att, x$att.egt) / c(x$overall.se, x$se.egt),
+      p.value   = 2 * (1 - stats::pnorm(abs(c(x$overall.att, x$att.egt) / c(x$overall.se, x$se.egt)))),
      conf.low  = c(x$overall.att - stats::qnorm(1 - x$DIDparams$alp/2) * x$overall.se, x$att.egt - x$crit.val.egt * x$se.egt),
      conf.high = c(x$overall.att + stats::qnorm(1 - x$DIDparams$alp/2) * x$overall.se, x$att.egt + x$crit.val.egt * x$se.egt),


In tidy.AGGTEobj for type == "group", the p.value expression has unbalanced parentheses, which will cause a parse error when the package is loaded. Adjust the parentheses so the 2 * (1 - pnorm(abs(...))) expression is syntactically valid (and consider assigning the statistic vector once to avoid repeating the division).

Copilot · 2026-03-02T19:36:39Z

tests/testthat/test-tidy.R

+test_that("tidy.MP returns expected columns", {
+  td <- broom::tidy(mp)
+  expect_s3_class(td, "data.frame")
+  expect_true(all(c("term", "group", "time", "estimate", "std.error",
+                    "statistic", "p.value",
+                    "conf.low", "conf.high",
+                    "point.conf.low", "point.conf.high") %in% names(td)))


These tests call broom::tidy(), but broom is not listed in DESCRIPTION under Suggests (or Imports). This will make R CMD check fail when broom isn't installed. Add broom to Suggests (recommended) or guard the tests with testthat::skip_if_not_installed("broom").

Copilot · 2026-03-02T19:36:39Z

tests/testthat/test-tidy.R

+
+#-----------------------------------------------------------------------------
+#
+# Tests for tidy(), glance(), and nobs() S3 methods


File header says these are tests for tidy(), glance(), and nobs(), but there are no glance() assertions in this file. Either add glance() coverage or update the header comment to avoid misleading future maintainers.

Suggested change

# Tests for tidy(), glance(), and nobs() S3 methods

# Tests for tidy() and nobs() S3 methods

Copilot · 2026-03-02T19:36:39Z

R/pre_process_did2.R

+  # Strip idname from clustervars in the caller's copy of args.
+  # validate_args works on a local copy, so we must repeat this here.
+  # Clustering at the unit level (idname) is already done implicitly.
+  if (!is.null(args$clustervars) && !is.null(args$idname) && (args$idname %in% args$clustervars)) {
+    args$clustervars <- setdiff(args$clustervars, args$idname)
+    if (length(args$clustervars) == 0L) args$clustervars <- NULL
+  }


The idname-stripping logic for clustervars is duplicated here and in validate_args(). Consider returning the (possibly modified) args from validate_args() (or factoring the stripping into a small helper) so the behavior stays consistent and future changes only need to be made in one place.

Copilot · 2026-03-02T19:36:40Z

man/nobs.MP.Rd

+Integer. The number of unique units in the data.
+}
+\description{
+Number of observations used to fit an MP object
+}


The title/description say "Number of observations", but the return value is "the number of unique units" (which can differ from row count in panel data). Consider updating the title/description to say "unique units" to avoid ambiguity about what nobs() reports for MP objects.

Copilot · 2026-03-02T19:36:40Z

man/nobs.AGGTEobj.Rd

+Integer. The number of unique units in the data.
+}
+\description{
+Number of observations used to fit an AGGTEobj object
+}


The title/description say "Number of observations", but the return value is "the number of unique units". Consider updating the title/description to explicitly say "unique units" to avoid confusion (especially for panel data where rows != units).

Copilot · 2026-03-02T19:36:40Z

DESCRIPTION

 Package: did
 Title: Treatment Effects with Multiple Periods and Groups
-Version: 2.3.1.902
+Version: 2.3.1.903


PR title/description references version 2.3.1.902, but DESCRIPTION bumps to 2.3.1.903. Please confirm the intended release/dev version so the PR metadata and package version stay in sync.

Copilot

Pull request overview

Copilot reviewed 11 out of 11 changed files in this pull request and generated 3 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-03-02T19:50:47Z

R/pre_process_did2.R

  # Flag for clustervars and weightsname
  checkvar_message <- "__ARG__ must be NULL or a character scalar that is a name of a column from the dataset."
  dreamerr::check_set_arg(args$weightsname, args$clustervars, "NULL | match", .choices = data_names, .message = checkvar_message, .up = 1)


In validate_args(), clustervars is passed to dreamerr::check_set_arg() which (per the comment) enforces scalar length. That means user inputs like clustervars = c(idname, extra1, extra2) will error with the generic dreamerr message before reaching the later custom length check (lines ~70-76). Consider checking the post-strip length of args$clustervars before calling check_set_arg (and then validating each element is in data_names), so users get the intended “only 1 extra cluster var” error and the later length check isn't dead/unreachable code.

Suggested change

# Flag for clustervars and weightsname

checkvar_message <- "__ARG__ must be NULL or a character scalar that is a name of a column from the dataset."

dreamerr::check_set_arg(args$weightsname, args$clustervars, "NULL | match", .choices = data_names, .message = checkvar_message, .up = 1)

# After stripping idname, ensure all remaining clustervars (if any) are column names.

if (!is.null(args$clustervars)) {

invalid_clust <- setdiff(args$clustervars, data_names)

if (length(invalid_clust) > 0L) {

stop("Each element of 'clustervars' must be the name of a column from the dataset.")

}

}

# Flag for clustervars and weightsname

checkvar_message <- "__ARG__ must be NULL or a character scalar that is a name of a column from the dataset."

dreamerr::check_set_arg(args$weightsname, "NULL | match", .choices = data_names, .message = checkvar_message, .up = 1)

Copilot · 2026-03-02T19:50:48Z

R/tidy.R

+#'   \item{statistic}{t-statistic (\code{estimate / std.error})}
+#'   \item{p.value}{two-sided pointwise p-value
+#'     (\code{2 * (1 - pnorm(abs(statistic)))}).


The documentation calls statistic a “t-statistic”, but the code (and docs for p.value) use a standard normal approximation via pnorm(). To avoid confusion for users (and to match broom conventions), consider renaming this to “z-statistic” or explicitly noting it is a normal-approximation test statistic.

Suggested change

#' \item{statistic}{t-statistic (\code{estimate / std.error})}

#' \item{p.value}{two-sided pointwise p-value

#' (\code{2 * (1 - pnorm(abs(statistic)))}).

#' \item{statistic}{z-statistic / normal-approximation test statistic

#' (\code{estimate / std.error})}

#' \item{p.value}{two-sided pointwise p-value based on the standard normal

#' approximation (\code{2 * (1 - pnorm(abs(statistic)))}).

Copilot · 2026-03-02T19:50:48Z

R/tidy.R

+#'   \item{term}{label for each estimate}
+#'   \item{estimate}{point estimate}
+#'   \item{std.error}{standard error}
+#'   \item{statistic}{t-statistic (\code{estimate / std.error})}


Same as above: the docs describe statistic as a “t-statistic”, but p-values are computed using pnorm() (normal approximation). Consider calling it a z-statistic (or clarifying it’s normal-approximation) to keep the return spec internally consistent.

Suggested change

#' \item{statistic}{t-statistic (\code{estimate / std.error})}

#' \item{statistic}{z-statistic (normal-approximation; \code{estimate / std.error})}

pedrohcgs added 6 commits March 2, 2026 09:52

Bump version to 2.3.1.903

40b7d53

pedrohcgs requested a review from Copilot March 2, 2026 18:33

Copilot started reviewing on behalf of pedrohcgs March 2, 2026 18:34 View session

pedrohcgs force-pushed the v2.3.1.902-bugfixes branch from cc29f67 to 40b7d53 Compare March 2, 2026 18:35

Copilot AI reviewed Mar 2, 2026

View reviewed changes

Address Copilot review: fix Rd markup, t-statistic wording, add tidy/…

60fbe21

…nobs tests

pedrohcgs requested a review from Copilot March 2, 2026 19:16

Copilot started reviewing on behalf of pedrohcgs March 2, 2026 19:16 View session

Copilot AI reviewed Mar 2, 2026

View reviewed changes

Address Copilot review: coerce nobs to integer, fix test warnings (bs…

960d9de

…trap/cband=FALSE)

pedrohcgs requested a review from Copilot March 2, 2026 19:26

Copilot started reviewing on behalf of pedrohcgs March 2, 2026 19:27 View session

Copilot AI reviewed Mar 2, 2026

View reviewed changes

Address Copilot review: nobs titles, broom in Suggests, fix test header

7288124

pedrohcgs requested a review from Copilot March 2, 2026 19:43

Copilot started reviewing on behalf of pedrohcgs March 2, 2026 19:44 View session

Copilot AI reviewed Mar 2, 2026

View reviewed changes

pedrohcgs merged commit 13f22ba into master Mar 2, 2026
10 checks passed

		#' \item{statistic}{t-statistic (\code{estimate / std.error})}
		#' \item{p.value}{two-sided pointwise p-value (\code{2*(1-pnorm(\|t\|))}).

	# Tests for tidy(), glance(), and nobs() S3 methods
	# Tests for tidy() and nobs() S3 methods

-  # Flag for clustervars and weightsname
-  checkvar_message <- "__ARG__ must be NULL or a character scalar that is a name of a column from the dataset."
-  dreamerr::check_set_arg(args$weightsname, args$clustervars, "NULL | match", .choices = data_names, .message = checkvar_message, .up = 1)
+  # After stripping idname, ensure all remaining clustervars (if any) are column names.
+  if (!is.null(args$clustervars)) {
+    invalid_clust <- setdiff(args$clustervars, data_names)
+    if (length(invalid_clust) > 0L) {
+      stop("Each element of 'clustervars' must be the name of a column from the dataset.")
+    }
+  }
+  # Flag for clustervars and weightsname
+  checkvar_message <- "__ARG__ must be NULL or a character scalar that is a name of a column from the dataset."
+  dreamerr::check_set_arg(args$weightsname, "NULL | match", .choices = data_names, .message = checkvar_message, .up = 1)

-#'   \item{statistic}{t-statistic (\code{estimate / std.error})}
-#'   \item{p.value}{two-sided pointwise p-value
-#'     (\code{2 * (1 - pnorm(abs(statistic)))}).
+#'   \item{statistic}{z-statistic / normal-approximation test statistic
+#'     (\code{estimate / std.error})}
+#'   \item{p.value}{two-sided pointwise p-value based on the standard normal
+#'     approximation (\code{2 * (1 - pnorm(abs(statistic)))}).

	#' \item{statistic}{t-statistic (\code{estimate / std.error})}
	#' \item{statistic}{z-statistic (normal-approximation; \code{estimate / std.error})}

Conversation

pedrohcgs commented Mar 2, 2026

Summary

Bug fixes

New broom methods

Documentation

JEL replication & diagnostics (pre-existing on branch)

Test plan

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 2, 2026

Choose a reason for hiding this comment

Uh oh!