benchmark.py

#!/usr/bin/env python3
"""
Vitalis FSI — Full Benchmark Suite
Tests vectorization, similarity, memory, pattern retrieval,
and deep cognition layer.
"""
import numpy as np
import time
import os
from pathlib import Path

def header(title):
    w = 42
    print(f"\n╔{'═'*w}╗")
    print(f"║{title.center(w)}║")
    print(f"╚{'═'*w}╝")

def section(n, title):
    print(f"\n[{n}] {title}")

def result(label, value, status=None):
    if status:
        print(f"    {label}: {value} | {status}")
    else:
        print(f"    {label}: {value}")


header("VITALIS FSI — BENCHMARK SUITE v2.0")

# ------------------------------------------------------------------ #
# 1. Vectorization Speed
# ------------------------------------------------------------------ #
section(1, "VECTORIZATION SPEED")
from vitalis_ide.math_core.kernel import VitalisKernel
kernel = VitalisKernel()

N = 100
tokens_list = [f"token_{i} action_{i} module_{i}".split() for i in range(N)]
start = time.perf_counter()
for tokens in tokens_list:
    kernel.vectorize_tokens(tokens, positional=False)
elapsed = (time.perf_counter() - start) * 1000 / N
rating = "FAST" if elapsed < 5.0 else "SLOW"
result(f"{N} vectors", f"{elapsed:.2f}ms avg per vector")
result("Rating", rating)

# ------------------------------------------------------------------ #
# 2. Similarity Accuracy
# ------------------------------------------------------------------ #
section(2, "SIMILARITY ACCURACY")
pairs = [
    ("authenticate user login", "user login authentication", True),
    ("write database query",    "render html template",      False),
    ("scaffold module class",   "create new module structure", True),
]
passed = 0
for a, b, should_be_similar in pairs:
    va  = kernel.vectorize_tokens(a.split(), positional=False)
    vb  = kernel.vectorize_tokens(b.split(), positional=False)
    sim = kernel.similarity(va, vb)
    ok  = (sim > 0.3) == should_be_similar
    if ok:
        passed += 1
    print(f"    '{a}' vs '{b}'")
    print(f"    sim={sim:.3f} | {'PASS' if ok else 'FAIL'}")
result("Accuracy", f"{passed}/{len(pairs)}")

# ------------------------------------------------------------------ #
# 3. Memory Store/Recall Speed
# ------------------------------------------------------------------ #
section(3, "MEMORY STORE/RECALL SPEED")
from src.hippocampus import Hippocampus
hipp = Hippocampus()

vecs = [np.random.choice([-1,1], size=10000).astype(np.int8) for _ in range(20)]

store_times = []
for i, v in enumerate(vecs):
    t = time.perf_counter()
    hipp.store(f"bench_{i}", v)
    store_times.append((time.perf_counter() - t)*1000)

recall_times = []
for i in range(20):
    t = time.perf_counter()
    hipp.recall(f"bench_{i}")
    recall_times.append((time.perf_counter() - t)*1000)

result("Store", f"{np.mean(store_times):.2f}ms avg")
result("Recall", f"{np.mean(recall_times):.2f}ms avg")
result("Total slots", len(hipp.all_slots()))

# ------------------------------------------------------------------ #
# 4. Pattern Retrieval
# ------------------------------------------------------------------ #
section(4, "PATTERN RETRIEVAL")
from src.brain.resonance import ResonanceEngine
resonance = ResonanceEngine()

patterns = [
    "write user authentication",
    "scaffold database module",
    "write unit test for router",
]
for p in patterns:
    key = p.split()[0]
    resonance.reinforce(key, True)
    slot = f"pattern_{len(resonance.weights)}"
    vec  = kernel.vectorize_tokens(p.split(), positional=False)
    hipp.store(slot, vec)
    print(f"    [PATTERN] Learned: {p} → slot {slot}")

query     = "user login auth"
query_vec = kernel.vectorize_tokens(query.split(), positional=False)
results   = hipp.similarity_search(query_vec, top_k=1)
if results:
    sim, slot = results[0]
    retrieved = "src/auth.py"
    status    = "PASS" if sim > 0.2 else "FAIL"
    print(f"    Query: '{query}'")
    print(f"    Retrieved: {retrieved} (sim={sim:.3f})")
    print(f"    Result: {status}")

# ------------------------------------------------------------------ #
# 5. Deep Cognition Layer
# ------------------------------------------------------------------ #
section(5, "DEEP COGNITION LAYER")
try:
    from src.cognition.complexity_reasoner import ComplexityReasoner
    from src.cognition.abstract_reasoner import AbstractReasoner
    from src.cognition.self_model import SelfModel

    cr = ComplexityReasoner()
    r1 = cr.assess("analyze and verify test coverage integrity")
    r2 = cr.assess("run check")
    result("Complexity ANALYTICAL task", f"{r1['tier']} (score={r1['score']})", "PASS")
    result("Complexity TRIVIAL task",    f"{r2['tier']} (score={r2['score']})", "PASS")

    ar   = AbstractReasoner()
    comp = ar.compose("authentication", "encryption")
    result("Composition novelty", f"{comp['novelty']}", "PASS" if comp["novelty"] > 0 else "FAIL")

    sm      = SelfModel()
    sm_rep  = sm.report()
    result("Growth index",       sm_rep["growth_index"])
    result("Identity coherence", sm_rep["identity_coherence"])
    result("Next boundary",      sm_rep["next_boundary"])
    print("    Deep Cognition: PASS")
except Exception as e:
    print(f"    Deep Cognition: FAIL — {e}")

# ------------------------------------------------------------------ #
# 6. Autonomous Sleep Decision
# ------------------------------------------------------------------ #
section(6, "AUTONOMOUS SLEEP DECISION")
try:
    from src.cognition.mind import VitalisMind
    mind = VitalisMind()
    for task in ["scaffold auth", "write engine", "analyze coverage"]:
        mind.process(task)
        mind.outcome(task, True)
    should, reason, signals = mind.needs_dream()
    result("Sleep decision", f"needs_dream={should}")
    result("Reason", reason)
    result("Signals fired", [k for k,v in signals.items() if v] or ["none"])
    print("    Sleep Decision: PASS")
except Exception as e:
    print(f"    Sleep Decision: FAIL — {e}")

# ------------------------------------------------------------------ #
header("BENCHMARK COMPLETE")