Issues

step_lag_difference()`/`step_growth_rate() `use` left_join` and cannot populate the synthetic forecast rows created by lag/ahead steps — empty predictions when all lags > 0

#515

· docxology opened

on Sep 16, 2026

step_climate()` aggregates the season containing each training row's target week — in-sample contamination in backtests

#514

· docxology opened

on Sep 16, 2026

pad_to_end()` (locf bake) crashes with an uninformative error when a selected column is all NA

#513

· docxology opened

on Sep 16, 2026

epi_recipe(reference_date =)` is stored but never consumed by any date machinery

#512

· docxology opened

on Sep 16, 2026

check_interminable_latency() `only fires for` method = "locf"`;` extend_ahead`/`extend_lags` never warn (and their tests expect it)

#511

· docxology opened

on Sep 16, 2026

Predict-time latency drift is never checked for proper `epi_df` inputs (the comparison runs only on the odd path)

#510

· docxology opened

on Sep 16, 2026

layer_residual_quantiles()` description recommends itself for distributional models

#509

· docxology opened

on Sep 16, 2026

R-CMD-check `and` pkgdown `workflows run arbitrary code without a minimal` permissions:` block

#508

· docxology opened

on Sep 16, 2026

Third-party GitHub Actions pinned by mutable tags, not commit SHAs; no dependabot config (plus checkout version drift in `pr-commands`)

#507

· docxology opened

on Sep 16, 2026

Scalar `fixed_latency` is recomputed from the data per column; the overwrite is a dead self-assignment

#506

· docxology opened

on Sep 16, 2026

Unpinned `Remotes` dependencies — including a personal-account fork — at floating HEAD, with a live API key in the CI check environment

#505

· docxology opened

on Sep 16, 2026

flatline_forecaster()`/`cdc_baseline_forecaster() `silently use data beyond a user-supplied earlier` forecast_date`; the three args constructors disagree on the same condition

#504

· docxology opened

on Sep 16, 2026

Navigation Menu

All issues

step_lag_difference()`/`step_growth_rate() `use` left_join` and cannot populate the synthetic forecast rows created by lag/ahead steps — empty predictions when all lags > 0

step_climate()` aggregates the season containing each training row's target week — in-sample contamination in backtests

pad_to_end()` (locf bake) crashes with an uninformative error when a selected column is all NA

epi_recipe(reference_date =)` is stored but never consumed by any date machinery

check_interminable_latency() `only fires for` method = "locf"`;` extend_ahead`/`extend_lags` never warn (and their tests expect it)

Predict-time latency drift is never checked for proper `epi_df` inputs (the comparison runs only on the odd path)

layer_residual_quantiles()` description recommends itself for distributional models

R-CMD-check `and` pkgdown `workflows run arbitrary code without a minimal` permissions:` block

Third-party GitHub Actions pinned by mutable tags, not commit SHAs; no dependabot config (plus checkout version drift in `pr-commands`)

Scalar `fixed_latency` is recomputed from the data per column; the overwrite is a dead self-assignment

Unpinned `Remotes` dependencies — including a personal-account fork — at floating HEAD, with a live API key in the CI check environment

flatline_forecaster()`/`cdc_baseline_forecaster() `silently use data beyond a user-supplied earlier` forecast_date`; the three args constructors disagree on the same condition

Search results