Rewrite as a probabilistic model with walk-forward evaluation.

Replace the 2024 model (model.py, ~2000 lines) with the btcmodel package, the
baseline for future work:

- Forecasts are quantiles of log price at each horizon, scored with CRPS in a
  walk-forward backtest (origins every 30 days from 2014, horizons 1 month to
  4 years). Skill is relative to a zero-drift random walk, with circular
  block-bootstrap intervals and a count of independent windows.
- Development data stops at 2024-11-26, the last day the 2024 model saw.
  Later outcomes are a holdout, scored only by `backtest --holdout`.
- Models: random_walk, drift_rw, and cycle (the 2024 model's cycle-position
  drift, now kernel-smoothed and recency-weighted). On development data
  nothing beats the random walk with confidence; cycle loses at every horizon.
- Prices: the Investing.com archive moves to data/ (cut at 2024-11-26; its
  last row was intraday) and is extended with Coinbase daily closes by
  `update`.

Also: Nix flake dev shell (Python 3.13, pandas 3), ruff in place of black,
pytest suite, and a rewritten README. NOTES.md is removed as inaccurate, and
poetry is dropped.

justfile

@@ -1,11 +1,34 @@
-run *name:
-    python3 ./model.py -n "{{ name }}"
+# List recipes
+default:
+    @just --list
+
+# Fetch new daily prices from Coinbase
+update:
+    python -m btcmodel update
+
+# Score models on development data (outcomes up to the dev cutoff)
+backtest *args:
+    python -m btcmodel {{ args }} backtest
+
+# Score models on held-out outcomes; run sparingly, never while tuning
+holdout *args:
+    python -m btcmodel {{ args }} backtest --holdout
+
+# Forecast from the latest price
+forecast *args:
+    python -m btcmodel {{ args }} forecast
+
+test:
+    pytest -q
 
 fmt:
-    black ./*.py
+    ruff format .
+    ruff check --fix --select I .
+    nix fmt flake.nix
 
 lint:
-    ruff check ./model.py
+    ruff check .
+    ruff format --check .
 
 clean:
     rm -rf output