Rerun Query

Python SDK to query and extract data from Rerun files.

Rerun Compatibility

This project is only compatible with Rerun 0.18. Please make sure the source rrd file is created with this version of Rerun SDK [Rust].

Note

The Blueprint of Rerun data (controls the layout of the viewer) is ignored for now, while the entity_paths were still included in the return value of list_entity_paths().

Please open an issue if you need this feature or other help.

Install

This project depends on numpy, so please make sure install it with this package together.

pip install numpy rerun-query

The retrieved data is in numpy arrays.

Usage

Use this package:

import requery

data = requery.query_data_entities(file_path, "", "")
for data_row in data:
    print(f"Entity Path: {data_row.entity_path}")
    for timeline_key, times in data_row.timelines.items():
        print(f"Timeline({timeline_key}) - {times}")
    data_object = data_row.data
    for index, data in enumerate(data_object[:10]):
        print(f"- {index + 1} {data}")

Example

You can find running example and sample data file in examples folder.

To run the example Python code:

cd ./examples
python3 -m venv .venv
source .venv/bin/activate
python3 main.py

The output includes entity paths, meta data, and tensor data in numpy arrays.

To view the logs in detail, run the program as:

RUST_LOG=debug python3 main.py

If the program crashes unexpectedly, try to diagnose with:

RUST_BACKTRACE=1 python3 main.py