feat(cpp/datascience): CPU stats + MCMC primitives

Nuevo dominio cpp/functions/datascience con primitivas puras CPU para post-
proceso de samples Monte Carlo y diagnostico de cadenas MCMC. Diseñadas como
gemelas CPU de los kernels GPU (rng pareja con gpu_rng_glsl, MH 1D/ND con
mc_metropolis_hastings_gpu) para validar numericamente y para datasets
pequeños donde el dispatch GPU no compensa.

- rng: xoshiro256++ con uniform / normal (Box-Muller) / below (Lemire) /
  categorical. Determinista bit-exacto dado seed.
- stats_summary: sum (Kahan), mean, var/std (Welford one-pass), min, max,
  quantile / percentile (R type-7).
- autocorr: r(k), ACF, tau_int (Sokal) — diagnostico ACF y ESS.
- rhat_ess: Gelman-Rubin clasico y split + ESS basico (multi-chain).
- beta_dist: lgamma (Lanczos), beta_pdf, beta_cdf (continued fraction),
  beta_quantile, mean/var/std — para inferencia Beta-Binomial.
- drawdown: max_dd absoluto/pct + underwater series para sesiones
  simuladas y backtests.
- samples_to_grid_2d: binning 2D CPU para alimentar heatmap_cpp_viz /
  contour_cpp_viz desde samples (x[], y[]).
- metropolis_hastings: MH 1D y ND con target log-pdf como std::function
  (no normalizada).

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

cpp/functions/datascience/drawdown.cpp

@@ -0,0 +1,48 @@
+#include "datascience/drawdown.h"
+
+namespace fn::ds {
+
+DrawdownResult drawdown_max(const double* equity, std::size_t n) {
+    DrawdownResult r{};
+    if (equity == nullptr || n == 0) return r;
+
+    double peak = equity[0];
+    std::size_t peak_i = 0;
+    double max_dd = 0.0;
+    std::size_t best_peak_i = 0;
+    std::size_t best_trough_i = 0;
+
+    for (std::size_t i = 0; i < n; ++i) {
+        double v = equity[i];
+        if (v > peak) {
+            peak = v;
+            peak_i = i;
+        } else {
+            double dd = peak - v;
+            if (dd > max_dd) {
+                max_dd = dd;
+                best_peak_i = peak_i;
+                best_trough_i = i;
+            }
+        }
+    }
+    r.max_dd = max_dd;
+    r.peak_idx = best_peak_i;
+    r.trough_idx = best_trough_i;
+    double peak_value = equity[best_peak_i];
+    if (peak_value > 0.0) {
+        r.max_dd_pct = max_dd / peak_value;
+    }
+    return r;
+}
+
+void drawdown_series(const double* equity, std::size_t n, double* out) {
+    if (equity == nullptr || out == nullptr || n == 0) return;
+    double peak = equity[0];
+    for (std::size_t i = 0; i < n; ++i) {
+        if (equity[i] > peak) peak = equity[i];
+        out[i] = peak - equity[i];
+    }
+}
+
+} // namespace fn::ds