datascope-ml 1.0.2

The official Python SDK for DataScope: The Machine Learning Observability Platform

DataScope ML SDK

The official Python SDK for DataScope, the premier Machine Learning Observability Platform.

Installation

pip install datascope-ml

Quick Start

Seamlessly upload and analyze your Pandas DataFrames directly from your Jupyter Notebook or Python IDE.

import pandas as pd
import datascope

# Load your local dataset
df = pd.read_csv("my_training_data.csv")

# Initialize the DataScope client
client = datascope.Client(api_key="YOUR_API_KEY")

# Upload and generate analysis
# This automatically opens your browser to the results dashboard!
client.analyze(
    df, 
    project_name="fraud_detection_v2", 
    target_column="is_fraud",
    prediction_type="classification"
)

Features

• Zero Friction: No need to export CSVs and drag-and-drop into a browser.
• Privacy First: Data dictionaries and outlier math can run locally, minimizing data sent over the network.
• Auto-Browser Launch: Automatically opens your Vercel-hosted dashboard as soon as the backend completes analysis.