Scoperta scientifica tramite similarità del coseno di Usai

import numpy as np

# Feature dimensions in archeological/anthropological multi-proxy space:
# Let's define the latent vector representation of Sardinian Ozieri / Sub-Ozieri / Bonnanaro
# and compare it with:
# 1. Corsica Terrinien / Fontanaccia
# 2. Southern France / Pyrenees / Fontbouisse / SOM (Chasseen, Treilles)
# 3. Sicily (San Cono - Piano Notaro / Malpasso / Castelluccio)
# 4. Malta (Saflieni / Tarxien)
# 5. Balearic (Talayotic / Pre-Talayotic navetas & caves)
# 6. Caucasus (Maikop / Dolmen culture - Zhane, Gelendzhik)
# 7. Anatolia (Lycian rock tombs)
# 8. Korea (Mumun dolmens)
# 9. Atlantic Megalithism (Iberia / Brittany - Los Millares, Gavrinis, Carnac)
# 10. Levant / Jordan (Golan dolmens)

# Dimensions:
# Dim 0: Chronological overlap with Ozieri (3500-2700 BC)
# Dim 1: Hypogeism (carving living rock vs open air megalith)
# Dim 2: Megalithic chamber (orthostats + capstone)
# Dim 3: Rock-cut domestic imitation (columen, beams, false door, hearth)
# Dim 4: Bovine/bucranium iconography & horned symbolism
# Dim 5: Ochre/pigment ritual use
# Dim 6: Obsidian trade source (Monte Arci / West Med vs Lipari vs Melos vs Anatolian/Caucasus)
# Dim 7: Ceramic style (tripod, carinated, Ozieri incision/graffiti)
# Dim 8: Metallurgy stage (Chalcolithic arsenical copper vs Copper flat axes vs Bronze Age alloys)
# Dim 9: aDNA affinity (EEF I2a1a1-M26 vs Steppe R1b vs CHG J2/L vs Asian)
# Dim 10: Maritime accessibility/Least Cost Path from Sardinia (Tirrenian basin proximity)

features = [
    "Ozieri_Overlap", "Hypogean_Carving", "Megalith_Trilith", "Domestic_Imitation",
    "Bucrania_Symbolism", "Ochre_Ritual", "Monte_Arci_Obsidian", "Ozieri_Ceramics",
    "Early_Copper_Tech", "aDNA_EEF_I2a", "Tyrrhenian_Proximity"
]

# Vectors normalized or weighted
# Sardinia (Ozieri/Domus de Janas reference)
v_sardinia = np.array([1.0, 1.0, 0.2, 1.0, 1.0, 1.0, 1.0, 1.0, 0.8, 1.0, 1.0])

# 1. Corsica (Terrinien / megalithism & hypogea)
v_corsica = np.array([0.9, 0.7, 0.8, 0.3, 0.2, 0.8, 0.9, 0.7, 0.7, 0.9, 0.95])

# 2. Sicily (San Cono-Piano Notaro / Grotta Zubbia)
v_sicily_neol = np.array([0.9, 0.9, 0.1, 0.4, 0.2, 0.9, 0.3, 0.6, 0.6, 0.8, 0.75])

# 3. Southern France (Chasseen / Treilles hypogea / Fontbouisse)
v_s_france = np.array([0.85, 0.8, 0.5, 0.2, 0.1, 0.8, 0.4, 0.5, 0.6, 0.75, 0.7])

# 4. Malta (Hal Saflieni / Tarxien)
v_malta = np.array([0.85, 1.0, 0.6, 0.7, 0.1, 1.0, 0.0, 0.4, 0.5, 0.7, 0.6])

# 5. Sa Coveccada (Sardinia Chalcolithic / Bonnanaro Dolmen)
v_sa_coveccada = np.array([0.4, 0.0, 1.0, 0.1, 0.2, 0.4, 0.6, 0.3, 0.9, 0.95, 1.0])

# 6. Atlantic Iberia / Los Millares
v_iberia_millares = np.array([0.8, 0.4, 0.9, 0.1, 0.2, 0.7, 0.0, 0.4, 0.8, 0.6, 0.5])

# 7. Paris Basin (SOM - Seine-Oise-Marne / Wartberg - Dalle perforée)
v_som_paris = np.array([0.7, 0.5, 0.9, 0.1, 0.0, 0.6, 0.0, 0.2, 0.5, 0.65, 0.3])

# 8. Caucasus (Gelendzhik / Zhane / Maikop)
v_caucasus = np.array([0.7, 0.1, 1.0, 0.0, 0.0, 0.3, 0.0, 0.1, 1.0, 0.0, 0.05])

# 9. Anatolia (Lycian rock-cut tombs - Myra)
v_lycia = np.array([0.0, 1.0, 0.1, 0.9, 0.0, 0.1, 0.0, 0.0, 0.0, 0.0, 0.1])

# 10. Korea (Gochang Dolmens)
v_korea = np.array([0.3, 0.0, 1.0, 0.0, 0.0, 0.1, 0.0, 0.0, 0.4, 0.0, 0.0])

sites = {
    "Corsica (Terrinien / Fontanaccia)": v_corsica,
    "Sicilia (San Cono - Piano Notaro / Zubbia)": v_sicily_neol,
    "Francia Meridionale (Chasséen / Ipogei di Fontvieille)": v_s_france,
    "Malta (Hal Saflieni)": v_malta,
    "Sardegna Dolmenica (Sa Coveccada)": v_sa_coveccada,
    "Iberia Sud-Orientale (Los Millares / Tholoi)": v_iberia_millares,
    "Bacino di Parigi (Cultura SOM / Allées couvertes)": v_som_paris,
    "Caucaso (Dolmen di Gelendzhik / Zhane)": v_caucasus,
    "Anatolia (Tombe Rupestri di Licia)": v_lycia,
    "Corea (Dolmen di Gochang)": v_korea
}

def cosine_sim(a, b):
    return np.dot(a, b) / (np.linalg.norm(a) * np.linalg.norm(b))

results = {}
for name, vec in sites.items():
    sim = cosine_sim(v_sardinia, vec)
    results[name] = sim

# Sort by similarity
sorted_res = sorted(results.items(), key=lambda x: x[1], reverse=True)
for name, sim in sorted_res:
    print(f"{name}: {sim:.4f}")

Sicilia (San Cono - Piano Notaro / Zubbia): 0.9313
Corsica (Terrinien / Fontanaccia): 0.9117
Francia Meridionale (Chasséen / Ipogei di Fontvieille): 0.8989
Malta (Hal Saflieni): 0.8574
Iberia Sud-Orientale (Los Millares / Tholoi): 0.7700
Sardegna Dolmenica (Sa Coveccada): 0.7352
Bacino di Parigi (Cultura SOM / Allées couvertes): 0.7071
Anatolia (Tombe Rupestri di Licia): 0.5023
Caucaso (Dolmen di Gelendzhik / Zhane): 0.4483
Corea (Dolmen di Gochang): 0.2634