Data Framework

The CaptureGraph Data Framework lets you load and analyze captured data in Python. It provides typed containers, vectorized access, and seamless integration with analysis tools.

Overview

While the Procedure DSL defines how to capture data, the Data Framework handles what to do with it afterward:

Procedure DSL → iOS App → Captured Data → Data Framework → Analysis

The framework:

• Loads captured data using the ProcedureManifest
• Converts files to typed Python objects (images, locations, etc.)
• Provides vectorized access across sessions
• Integrates with NumPy and Pandas

Quick Start

import capturegraph.data as cg

# Load a captured target
target = cg.CaptureTarget("/path/to/MyTarget")

# Access sessions
sessions = target.sessions.daily_capture
print(f"Found {len(sessions)} sessions")

# Vectorized access across all sessions
all_photos = sessions.photo
all_notes = sessions.notes

# Work with data
for session in sessions:
    print(f"{session.date}: {session.notes}")

Section Contents

• Loading Captures

Load targets, traverse manifests, and access session data

• Containers

List and Dict containers with vectorized operations and NumPy integration

• Advanced Operations

Vectorized functions, parallel processing, and data reshaping utilities