Configurare l'ambiente Python

Perché gli ambienti virtuali sono importanti

# Create and activate a virtual environment
# Run these commands in your terminal

python3 -m venv .venv          # create environment in .venv/ folder
source .venv/bin/activate      # activate on macOS/Linux
# .venv\Scripts\activate       # activate on Windows

python --version               # verify you are using the right Python

Installare l'SDK Python di OpenAI

# Install the OpenAI SDK
pip install openai

# For production projects, pin the version:
pip install 'openai>=1.30.0,<2.0.0'

# Save your dependencies to requirements.txt:
pip freeze > requirements.txt

# Install from requirements.txt on a new machine:
pip install -r requirements.txt

Ottenere e memorizzare la chiave API

# .env file (never commit this!)
# OPENAI_API_KEY=sk-proj-...

# In your Python script:
import os
from dotenv import load_dotenv

load_dotenv()  # reads .env and sets environment variables

api_key = os.environ.get('OPENAI_API_KEY')
if not api_key:
    raise ValueError('OPENAI_API_KEY environment variable not set!')

print('API key loaded:', api_key[:8] + '...')  # only print prefix

Inizializzare il client OpenAI

from openai import OpenAI

# API key is read from OPENAI_API_KEY environment variable automatically
client = OpenAI()

# Or explicitly:
# client = OpenAI(api_key='sk-proj-...')  # only for quick tests

# For alternative providers:
# client = OpenAI(
#     api_key='your-together-key',
#     base_url='https://api.together.xyz/v1'
# )

Il primo controllo dello stato dell'API

from openai import OpenAI

client = OpenAI()

# Simple health check: list available models
try:
    models = client.models.list()
    print(f'Connection successful! Found {len(list(models))} models.')
except Exception as e:
    print(f'Connection failed: {e}')

Buone pratiche per la struttura dei progetti

# Recommended project structure:
# my_ai_project/
# |-- .env              (secrets, gitignored)
# |-- .env.example      (template, committed)
# |-- .gitignore
# |-- requirements.txt
# |-- README.md
# |-- src/
# |   |-- __init__.py
# |   |-- client.py    (OpenAI client initialization)
# |   |-- prompts.py   (prompt templates)
# |   |-- main.py
# |-- tests/
# |-- notebooks/
# |-- Makefile

Gestire più chiavi API in modo sicuro

# .env.example - commit this as a template
# OPENAI_API_KEY=sk-proj-...your key here...
# OPENAI_ORG_ID=org-...optional org id...
# COHERE_API_KEY=...cohere key...
# PINECONE_API_KEY=...pinecone key...
# PINECONE_ENVIRONMENT=us-east-1-aws

# In code, load all at once:
from dotenv import load_dotenv
import os

load_dotenv()

config = {
    'openai_key': os.environ['OPENAI_API_KEY'],
    'cohere_key': os.environ.get('COHERE_API_KEY'),  # optional
}

Usare i gestori delle versioni di Python

# Install pyenv (macOS via Homebrew)
# brew install pyenv

# Install Python 3.11
# pyenv install 3.11.9

# Set Python version for this project directory
# pyenv local 3.11.9

# This creates a .python-version file:
# cat .python-version
# 3.11.9

# Now python3 automatically uses 3.11.9 in this directory

Testare la configurazione con un prompt reale

from openai import OpenAI
from dotenv import load_dotenv

load_dotenv()
client = OpenAI()

response = client.chat.completions.create(
    model='gpt-4o-mini',
    messages=[
        {'role': 'system', 'content': 'You are a test responder.'},
        {'role': 'user', 'content': 'Reply with just the word READY.'}
    ],
    temperature=0,
    max_tokens=5
)

print('Response:', response.choices[0].message.content)
print('Model used:', response.model)
print('Tokens used:', response.usage.total_tokens)

Usare uv per una gestione più rapida dei pacchetti

# Install uv (macOS/Linux)
# curl -LsSf https://astral.sh/uv/install.sh | sh

# Create a project with uv
# uv init my-ai-project
# cd my-ai-project

# Add dependencies
# uv add openai python-dotenv

# Run your script using the uv-managed environment
# uv run python main.py

# Generate requirements.txt for compatibility
# uv pip compile pyproject.toml -o requirements.txt

Le variabili d'ambiente in produzione

# Centralized config loader pattern for production
import os
from dataclasses import dataclass

@dataclass
class AppConfig:
    openai_api_key: str
    environment: str
    log_level: str

    @classmethod
    def from_env(cls):
        key = os.environ.get('OPENAI_API_KEY')
        if not key:
            raise EnvironmentError('OPENAI_API_KEY must be set')
        return cls(
            openai_api_key=key,
            environment=os.environ.get('ENVIRONMENT', 'development'),
            log_level=os.environ.get('LOG_LEVEL', 'INFO')
        )

config = AppConfig.from_env()

Verifica rapida

Riepilogo della lezione