Py-DS Library
The py-ds library provides a Python client for interacting with DeepChem Server. It offers a convenient, programmatic way to upload data, run featurization tasks, and manage datasets without dealing with raw HTTP requests.
Overview
The py-ds library simplifies common workflows by providing:
High-level functions for data upload and featurization
Error handling and retry logic for robust operations
Batch processing capabilities for handling multiple datasets
Integration helpers for popular data science tools
Installation
Install the py-ds library using pip:
pip install py-ds
Or install from source:
cd py-ds
pip install -e .
Configuration
Configure the library to connect to your DeepChem Server using environment variables, configuration files, or programmatically. For detailed configuration options, refer to the module documentation.
Interactive API Testing
For the most up-to-date examples and interactive testing of the underlying API that the py-ds library uses, visit http://localhost:8000/docs when your DeepChem Server is running.
The Swagger UI provides:
Live endpoint testing: Test API calls interactively
Request/response examples: See real data formats
Parameter documentation: Understand all available options
Schema definitions: Review data structures
Library Documentation
For detailed documentation of the py-ds library functions and classes, refer to the auto-generated documentation from the source code docstrings.
Getting Started
Install the library: Follow the installation instructions above
Configure connection: Set up connection to your DeepChem Server
Explore the API: Visit http://localhost:8000/docs to understand available endpoints
Check examples: Review the examples section for common workflows
Read the source: Examine the library source code for detailed implementation
For comprehensive examples, testing, and up-to-date API information, always refer to:
Interactive docs: http://localhost:8000/docs
Source code: Check the py-ds directory for implementation details
Module documentation: Auto-generated docs from docstrings