File size: 1,431 Bytes
a6f7c7a | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 | """
example.py
Minimal usage example for hv-temporal-context.
"""
from hv_temporal_context import HVTemporalContext
def main():
model = HVTemporalContext.load("temporal_context_model")
queries = [
("what is the forecast for rome", "weather"),
("how do i debug a python function", "coding"),
("how much did i spend on groceries", "finance"),
("how do i cook soup", "cooking"),
]
print("Pure similarity retrieval:")
print()
for q, _ in queries:
idx, scores = model.retrieve(q, k=3)
print(f" {q}")
for i, (j, s) in enumerate(zip(idx, scores)):
print(f" top{i+1} [{model.topics[j]:<10}] sim={s:.3f}")
print()
print("Position-conditional retrieval (with topic hint):")
print()
for q, topic in queries:
idx, scores = model.retrieve_temporal(q, topic, k=3)
print(f" {q} (hint: {topic})")
for i, (j, s) in enumerate(zip(idx, scores)):
print(f" top{i+1} [{model.topics[j]:<10}] sim={s:.3f}")
print()
print("Topic-shift detection:")
print()
boundaries = model.topic_boundaries()
print(f" detected boundaries at positions: {boundaries[:10]}")
print(f" total boundaries: {len(boundaries)}")
print(f" median surprise magnitude: "
f"{float(__import__('numpy').median(model.surprise_mags)):.3f}")
if __name__ == "__main__":
main() |