Skip to main content

​
Code Preparation

​
Dependencies

Clearly specify your Python dependencies to ensure consistent execution environments: 
# requirements.txt
pandas>=1.5.0
numpy>=1.24.0
scikit-learn>=1.2.0
If no requirements.txt is provided we will attempt to automatically detect any imports that could be needed during execution. However, this can be unreliable and directly stating requirements is encouraged.

​
File Paths

Use paths relative to your current working directory to ensure your code works both locally and on the cloud: 
from pathlib import Path

# Good: Relative paths
data_path = Path("data/input.csv")
output_path = Path("outputs/results.json")

# Avoid: Absolute paths
# data_path = "/home/user/project/data/input.csv"

​
Error Handling

Include proper error handling for robust execution: 
import logging

logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

try:
    # Your processing code
    result = process_data()
    logger.info("Processing completed successfully")
except Exception as e:
    logger.error(f"Processing failed: {e}")
    raise

​
Execution Optimization

​
Progress Monitoring

Provide feedback during long-running operations: 
from tqdm import tqdm

for item in tqdm(data, desc="Processing"):
    process_item(item)

​
Checkpointing

Save intermediate results for long-running processes: 
import pickle

# Save checkpoint
with open('checkpoint.pkl', 'wb') as f:
    pickle.dump(state, f)

# Resume from checkpoint
with open('checkpoint.pkl', 'rb') as f:
    state = pickle.load(f)

​
Using the API

​
Authentication

Always store credentials securely: 
import os

# Use environment variables
api_token = os.environ.get('LYCEUM_API_TOKEN')

# Never hardcode credentials in your code

​
Error Handling

Handle API responses properly: 
import requests

response = requests.post(
    'https://api.lyceum.technology/api/v2/external/execution/streaming/start',
    headers={'Authorization': f'Bearer {token}', 'Content-Type': 'application/json'},
    json={'code': code, 'nbcode': 0, 'execution_type': 'cpu'}
)

if response.status_code == 200:
    result = response.json()
else:
    print(f"Error: {response.status_code} - {response.text}")

​
Troubleshooting

​
Common Issues

Execution Fails to Start
  • Check your authentication token is valid
  • Verify you have sufficient credits
  • Ensure your code syntax is correct
Files Not Found
  • Verify files are uploaded to cloud storage
  • Check file paths are correct
  • Use the storage API to list available files
Out of Memory Errors
  • Process data in smaller chunks
  • Choose a machine type with more memory
  • Optimize your code to use less memory

​
Getting Help

  • Check execution logs for detailed error messages
  • Contact support at [email protected]
  • Review the API documentation for endpoint details
These best practices will help you use Lyceum Cloud effectively. For specific API endpoints and parameters, refer to the API Reference section.