Archive planning files for #20 + #43

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Author: NewGraphEnvironment

planning/archive/2026-05-issues-20-43-cd-compare-defaults-pvalue/.gitkeep

@@ -0,0 +1,29 @@
+## Outcome
+
+Two tightly coupled `cd_compare()` improvements landed together on
+one branch (issue #20 explicitly says "should land together" with
+#43). #20 wired the de-facto vignette framing
+(`window_a = 2015:2025`, `window_b = 1951:1980` — recent decade vs
+WMO-style standard normal) into the function as defaults so first-time
+users get a sensible answer with `cd::cd_compare(ts)`. #43 added a
+`test = "t"` argument (default) that emits a `p_value` column via
+Welch's two-sample t-test on the annual values within each window;
+`test = "wilcox"` switches to Mann-Whitney U; `test = NULL` skips and
+drops the column. Rows where either window has < 8 non-NA years get
+NA + a single batched warning naming affected variable/period rows.
+Both regional vignettes now show bare `cd::cd_compare(ts)` recipes
+and report both p-values side-by-side (`Δ p (windows)` from
+cd_compare + `Trend p (75-yr)` from cd_trend's MK), framed in
+narrative as different-question tests. Mid-PR the user flagged that
+the cumulative-impact section ("Recent vs Pre-warming") was buried
+~300 lines down between Snowpack and Spatial Pattern with a
+methodological-sounding heading; reordered it to come directly after
+Trends and renamed to "Recent Decade vs Pre-Warming Reference" so
+the headline +2 °C number lands before DTR/Snow/Spatial drill-down.
+Compare-table chunk recomputes `cmp` from cached `ts` to flow the
+new column through without re-running the heavy precompute scripts;
+`data-raw/<vignette>_vignette_data.R` updated for next refresh.
+181/0 tests, lint clean, both vignettes render.
+
+Closed by: PR #73 (squash `338c7e5`) → release commit `6bdd64b
+Release v0.3.0` → tag `v0.3.0`.