Skip to main content

Python API

The datagrok-api Python library is designed for integration with the Datagrok platform via REST API. It provides a convenient interface for working with data, tables, files, and other Datagrok objects.

Installation​

The library can be installed using pip:

pip install datagrok-api

Usage​

from datagrok_api import DatagrokClient

# Initialize with your server URL and API token
with DatagrokClient("https://public.datagrok.ai/api", "Bearer <your-token>") as grok:
    me = grok.users.current()
    print(f"Logged in as: {me.first_name} {me.last_name}")

API structure​

The Python API is built around a central DatagrokClient and a set of dedicated resource clients. DatagrokClient is the main entry point to the platform: it handles authentication and HTTP communication and exposes resource-specific clients as its properties.

Each resource module implements a dedicated client responsible for a specific domain, including user-related operations (users), group search and management (groups), upload and download of tables as Pandas DataFrames (tables), file upload and download (files), execution of server-side functions (functions), data connection management (connections), and shared entities and permissions (shares).

Typical usage follows this pattern:

grok.users.current()
grok.tables.download("JohnDoe:MyTable")
grok.groups.find("Chemists")

The models package contains typed representations of Datagrok entities (such as User, Group, DataConnection, Query, and Script) that are used as inputs and return values by the resource clients. All models share a common base class defined in model.py.