Power BI API Client

A lightweight Python client for interacting with the Power BI REST API. The library provides simple, object-oriented helpers to authenticate, list workspaces, fetch datasets and inspect data models.

Installation

Create a virtual environment and install the dependencies:

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Configuration

Authentication details are read from environment variables. Create a .env file or export the following variables in your environment:

POWERBI_TENANT_ID=<your-tenant-id>
POWERBI_CLIENT_ID=<your-client-id>
POWERBI_CLIENT_SECRET=<your-client-secret>

The client authenticates using an Azure AD service principal via the msal library.

Example Usage

Run the example script to list workspaces and datasets:

python examples/example_usage.py

You can specify a workspace to list datasets from using the --workspace-id flag. You can also print tables for a specific dataset using --dataset-id and display dashboards with --show-dashboards:

python examples/example_usage.py --workspace-id YOUR_WORKSPACE_ID \
    --dataset-id YOUR_DATASET_ID --show-dashboards

This will output a table of your Power BI workspaces and datasets using pandas DataFrames.

All get_raw methods accept a fetch_all flag. When set to True, the client automatically handles $top/$skip pagination and returns all records.

Extending

The codebase is structured to be easily extended for additional Power BI endpoints. Each entity class contains two core methods:

• get_raw - fetch raw JSON data from the API.
• to_dataframe - return the data as a pandas DataFrame for analysis.

Additional helpers and endpoints can be implemented following the same pattern.

Saving Output

Use powerbi_api_client.utils.save_json to write any Python object to a JSON file. The helper creates a json_data directory by default and accepts a custom file name:

from powerbi_api_client.utils import save_json

data = {"hello": "world"}
save_json(data, "example.json")