SniperGold_ML/ml/p3/baseline/train_small_mlp.py

# -*- coding: utf-8 -*-
"""P3-S.18 — SMALL MLP BASELINE.

Small MLP: input -> 16 hidden (tanh) -> 1 output (sigmoid), L2 alpha=1e-3,
max_iter=2000, fixed seed. NOT a reproduction of the legacy 19->12->2 MLP
(target differs: setup-outcome TP-before-SL, not 24-bar drift). Standardizer
fit on train only. Binary target: WIN vs LOSS (leads only).
"""
import csv
import json
import os
import sys

import numpy as np
from sklearn.neural_network import MLPClassifier
from sklearn.preprocessing import StandardScaler

HERE = os.path.dirname(os.path.abspath(__file__))
OUT = os.path.join(HERE, "output")
sys.path.insert(0, HERE)
sys.path.insert(0, os.path.normpath(os.path.join(HERE, "..", "setup_dataset")))

import prepare_dataset as PD          # noqa: E402

SEED = 42


def main():
    os.makedirs(OUT, exist_ok=True)
    ctx = PD.load_verified_population()
    rows = PD.build_rows(ctx)
    bin_rows = [r for r in rows if r["lead"] and r["outcome"] in PD.BINARY_CLASSES]
    bin_rows.sort(key=lambda r: (r["creation_bar"], r["setup_id"]))
    tr, va, te, _ = PD.temporal_split(rows)
    split_map = {}
    for part, tag in ((tr, "train"), (va, "val"), (te, "test")):
        for r in part:
            split_map[r["setup_id"]] = tag
    X = np.asarray([[r["feature_" + k] for k in PD.FEATURE_COLS]
                    for r in bin_rows], dtype=float)
    y = np.asarray([1.0 if r["outcome"] == "WIN" else 0.0
                    for r in bin_rows], dtype=float)
    splits = np.asarray([split_map[r["setup_id"]] for r in bin_rows])
    scaler = StandardScaler().fit(X[splits == "train"])
    Xs = scaler.transform(X)

    clf = MLPClassifier(hidden_layer_sizes=(16,), activation="tanh",
                        alpha=1e-3, max_iter=2000, random_state=SEED)
    clf.fit(Xs[splits == "train"], y[splits == "train"])
    p = clf.predict_proba(Xs)[:, 1]

    rows_out = []
    for i, r in enumerate(bin_rows):
        rows_out.append({"setup_id": r["setup_id"], "split": splits[i],
                         "outcome": r["outcome"], "y": int(y[i]),
                         "pred_mlp16": float(p[i])})
    with open(os.path.join(OUT, "p3_s18_predictions_mlp.csv"), "w",
              newline="", encoding="utf-8") as f:
        w = csv.DictWriter(f, fieldnames=list(rows_out[0].keys()))
        w.writeheader()
        for r in rows_out:
            w.writerow(r)
    with open(os.path.join(OUT, "p3_s18_mlp_model.json"), "w",
              encoding="utf-8") as f:
        json.dump({
            "model": "MLP(hidden=(16,),tanh,alpha=1e-3,max_iter=2000)",
            "feature_sha16": PD.make_manifest(rows, tr, va, te,
                                              True)["feature_sha16"],
            "seed": SEED, "n_iter": int(clf.n_iter_),
            "loss_curve_len": int(len(clf.loss_curve_)),
        }, f, indent=2)
    print("[saved] p3_s18_predictions_mlp.csv / mlp_model.json")
    return 0


if __name__ == "__main__":
    sys.exit(main())