# Suppress any warning messages for a cleaner experience
import warnings; warnings.filterwarnings("ignore")
import numpy as np
np.random.seed(42)

# This keeps our notebook tidy and focused.

# Let us import the key libraries
from openai import OpenAI
import os

# OpenAI is the main package for talking to the API.
# os lets us access environment variables safely.

# Set up your API key securely!
API_key = os.environ["OPENAI_API_KEY"]
client = OpenAI(api_key=API_key)

# This gives us a client we can use for all OpenAI requests.

# Let us try our first embedding call!
texts = [
    "OpenAI makes AI easy.",
    "Machine learning is fun",
    "Bananas are yellow."
]

embed = client.embeddings.create(
    model='text-embedding-3-small',
    input=texts
)

print("Number of texts embedded:", len(embed.data))

# What does an embedding look like?
embedding_vec = embed.data[0].embedding
print("First embedding vector (truncated):", embedding_vec[:10])

# Embeddings are lists of floats, which are just numbers.

# Simple cosine similarity implementation
import math

def cosine_similarity(vec1, vec2):
    dot = sum(a * b for a, b in zip(vec1, vec2))
    norm1 = math.sqrt(sum(a * a for a in vec1))
    norm2 = math.sqrt(sum(b * b for b in vec2))
    return dot / (norm1 * norm2)

# Let us compute similarities between our texts
sim_0_1 = cosine_similarity(embed.data[0].embedding, embed.data[1].embedding)
sim_0_2 = cosine_similarity(embed.data[0].embedding, embed.data[2].embedding)

print("First vs Second text:", round(sim_0_1, 2))
print("First vs Third text:", round(sim_0_2, 2))

# Close scores mean the texts are related.

# Let us handle possible errors with our embedding call
try:
    response = client.embeddings.create(
        model='text-embedding-3-small',
        input=["test"]
    )
    print("Success! Got an embedding.")
except Exception as e:
    print("There was an error:", e)

# Take a search phrase from the user and find the best matching sentence
search = input("Type your search phrase: ")
search_embed = client.embeddings.create(
    model='text-embedding-3-small',
    input=[search]
)
sims = [cosine_similarity(search_embed.data[0].embedding, emb.embedding) for emb in embed.data]
best_idx = sims.index(max(sims))
print("Best match:", texts[best_idx])

# Quick troubleshooting tips if your code is not working
print("Check these if you get errors:")
print("- Is your OPENAI_API_KEY set in the environment?")
print("- Is your network connection active?")
print("- Did you spell model and function names correctly?")