scripts/smoke_test.py

#!/usr/bin/env python3
"""Quick smoke test: build_data + train_ranker with 100k sample to verify bugfix."""
import os, sys, warnings, subprocess
warnings.filterwarnings("ignore")
sys.path.insert(0, "/tmp")
from huggingface_hub import HfApi, hf_hub_download

DS = "cedwyh/jinjing-shared-data"
hf_token = os.environ.get("HF_TOKEN")

# Step 1: Download fixed scripts
bd = hf_hub_download(DS, "build_data.py", repo_type="dataset")
tr = hf_hub_download(DS, "scripts/train_ranker.py", repo_type="dataset")

# Step 2: Run build_data with --sample 100000
print("=" * 60)
print("SMOKE TEST: build_data (100k sample)")
print("=" * 60)
r1 = subprocess.run(
    [sys.executable, bd, "--output", "/tmp/ranking_smoke.parquet",
     "--dataset", DS, "--no-use-priors", "--sample", "100000"],
    capture_output=True, text=True, timeout=300
)
print(r1.stdout)
if r1.returncode != 0:
    print(f"❌ FAILED: {r1.stderr[-500:]}")
    sys.exit(1)

# Verify columns include V5 features
import pandas as pd
df = pd.read_parquet("/tmp/ranking_smoke.parquet")
print(f"\n✅ Build data: {len(df):,} rows x {len(df.columns)} cols")
v5 = [c for c in ['buy_decay','sell_decay','zs_pos','zs_width','zs_time',
                   'momentum_div','volume_div','slope','trend_consistency',
                   'regime_prob','leg_progress','structure_progress'] if c in df.columns]
print(f"V5 features present: {len(v5)}/12")
ret = [c for c in ['ret_1d','ret_5d','ret_10d','ret_20d','ret_30d','ret_60d'] if c in df.columns]
print(f"ret features present: {len(ret)}/6")
zs = [c for c in ['zs_pos','zs_width','zs_time'] if c in df.columns]
print(f"zs features present: {len(zs)}/3")
print(f"Total columns: {len(df.columns)}")
print(f"Features: {[c for c in df.columns if c not in ['date','symbol','query_group','label_rank']]}")
del df

# Step 3: Verify NON_FEATURE_COLS fix
with open(tr) as f:
    content = f.read()
assert 'zs_width' not in content or 'zs_width' in content and 'Dead' not in content, "train_ranker still excludes zs!"
assert 'ret_1d' not in content or 'ret_1d' in content and 'Forward' not in content, "train_ranker still excludes ret!"
print("\n✅ train_ranker.py NON_FEATURE_COLS fix verified")

# Step 4: Patch target col and run train_ranker
tr_patched = "/tmp/tr_smoke.py"
with open(tr) as f:
    c = f.read()
c = c.replace('TARGET_COL = "label"', 'TARGET_COL = "label_rank"')
with open(tr_patched, "w") as f:
    f.write(c)

print("\n" + "=" * 60)
print("SMOKE TEST: train_ranker (100k sample, 1 window)")
print("=" * 60)
r2 = subprocess.run(
    [sys.executable, tr_patched, "--data", "/tmp/ranking_smoke.parquet",
     "--output", "/tmp/smoke_ranker"],
    capture_output=True, text=True, timeout=300
)
print(r2.stdout[-1000:])
if r2.returncode != 0:
    print(f"❌ FAILED: {r2.stderr[-500:]}")
    sys.exit(1)

print("\n" + "=" * 60)
print("✅ SMOKE TEST PASSED — All bugfixes verified")
print("=" * 60)