Python Examples

# Install the Tenzro Python SDK
pip install tenzro

# Or with async support
pip install tenzro[async]

# Basic setup
import tenzro
from tenzro import Tenzro

# Initialize client
client = Tenzro(api_key="sk_your_key_here")

# Async client
import asyncio
from tenzro import AsyncTenzro

async_client = AsyncTenzro(api_key="sk_your_key_here")

# Environment variable setup
import os
os.environ["TENZRO_API_KEY"] = "sk_your_key_here"

# Client will automatically use environment variable
client = Tenzro()

# Example usage
def main():
    try:
        response = client.cortex.generate(
            prompt="Hello, Tenzro!",
            model="gpt-4o"
        )
        print(f"Generated text: {response.content}")
    except Exception as error:
        print(f"Error: {error}")

if __name__ == "__main__":
    main()

# Async example
async def async_main():
    try:
        response = await async_client.cortex.generate(
            prompt="Hello, async Tenzro!",
            model="gpt-4o"
        )
        print(f"Generated text: {response.content}")
    finally:
        await async_client.close()

# Run async example
# asyncio.run(async_main())

import tenzro
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split
from sklearn.ensemble import RandomForestClassifier

client = Tenzro(api_key="sk_your_key_here")

# AI-powered data analysis
def analyze_dataset_with_ai(df):
    """Use AI to analyze a pandas DataFrame"""
    
    # Get basic statistics
    stats = df.describe().to_string()
    missing_data = df.isnull().sum().to_string()
    
    # Generate AI insights
    prompt = f"""
    Analyze this dataset and provide insights:
    
    Dataset shape: {df.shape}
    
    Statistics:
    {stats}
    
    Missing data:
    {missing_data}
    
    Columns: {list(df.columns)}
    
    Please provide:
    1. Key patterns and trends
    2. Data quality assessment
    3. Recommendations for further analysis
    4. Potential modeling approaches
    """
    
    response = client.cortex.generate(
        prompt=prompt,
        model="claude-3.5-sonnet",
        max_tokens=2000
    )
    
    return response.content

# Generate synthetic data with AI
def generate_synthetic_data(description, num_samples=1000):
    """Generate synthetic data based on description"""
    
    prompt = f"""
    Generate Python code to create a synthetic dataset with {num_samples} samples.
    
    Description: {description}
    
    Return only the Python code that creates a pandas DataFrame named 'df'.
    Use realistic data patterns and include some noise.
    """
    
    response = client.cortex.generate(
        prompt=prompt,
        model="gpt-4o",
        max_tokens=1000
    )
    
    # Execute the generated code
    exec(response.content)
    return locals().get('df')

# AI-assisted feature engineering
def suggest_features(df, target_column):
    """Get AI suggestions for feature engineering"""
    
    sample_data = df.head(10).to_string()
    
    prompt = f"""
    Given this dataset sample with target column '{target_column}':
    
    {sample_data}
    
    Suggest 5 feature engineering techniques that could improve model performance.
    Focus on:
    1. Creating interaction features
    2. Transformations
    3. Derived metrics
    4. Handling categorical variables
    5. Time-based features (if applicable)
    
    Provide specific Python code examples.
    """
    
    response = client.cortex.generate(
        prompt=prompt,
        model="claude-3.5-sonnet",
        max_tokens=1500
    )
    
    return response.content

# Example workflow
if __name__ == "__main__":
    # Load or generate data
    try:
        df = pd.read_csv("your_data.csv")
    except FileNotFoundError:
        # Generate synthetic data if file doesn't exist
        df = generate_synthetic_data(
            "E-commerce customer data with purchase history, demographics, and churn labels"
        )
    
    # AI-powered analysis
    insights = analyze_dataset_with_ai(df)
    print("AI Dataset Analysis:")
    print(insights)
    
    # Feature engineering suggestions
    if 'target' in df.columns:
        feature_suggestions = suggest_features(df, 'target')
        print("\nFeature Engineering Suggestions:")
        print(feature_suggestions)