ZoomingOnline is a web-based data visualization supporting the efficient and scalable Zarr format. Main use case is visualization of the waveform data from oscilloscopes, taken in segment mode.
Click on the links below to view the sample datasets, or copy the data file URLs to use in your own instance:
View 1nA dataset in ZoomingOnline
Data file URL (copy to use with your own instance):
https://s3.cloud.cyfronet.pl/zooming-online/1nA/1nA.zarr
View 64nA dataset in ZoomingOnline
Data file URL (copy to use with your own instance):
https://s3.cloud.cyfronet.pl/zooming-online/64nA/64nA.zarr
Additional file available on S3 (copy to use with ZoomingOnline):
https://s3.cloud.cyfronet.pl/zooming-online/20231204m4/20231204m4.zarr
The aim of ZoomingOnline is to provide a high-performance web viewer for interactively exploring massive, gigabyte-scale time-series datasets directly in a browser.
Key features include:
- Efficient Zooming: Fetches only the required high-resolution data chunks for the selected region, with a memory-efficient cache to prevent redundant downloads during interaction.
- Time-Based Controls: Features a three-level inset zoom interface with draggable windows and sliders that snap to meaningful time units (e.g., 10 µs, 100 ns) for precise analysis.
- Instant Overview: Utilizes pre-computed data pyramids within the Zarr format to load an initial overview of the entire dataset almost instantly.
- Data Toolkit: Includes a suite of CLI scripts for generating test data, converting from HDF5, and managing cloud storage.
| Script | Description |
|---|---|
generate_data.py |
Generate .zarr or .h5 oscilloscope-like datasets |
convert_hdf5_to_zarr.py |
Convert HDF5 files to Zarr format |
data_to_s3_importer.py |
Recursively convert + upload to S3 bucket |
cors_server.py |
Serve .zarr files locally with CORS enabled HTTP |
📖 See scripts.md for detailed CLI usage and arguments.
- Clone the repository
- Follow setup.md to install dependencies and configure
- Create or serve a Zarr dataset:
python src/generate_data.py -o waveform.zarr
python src/cors_server.py
- View in browser: http://localhost:8000/website, website/index.html or online at ZoomingOnline
- In the input field enter http://localhost:8000/waveform.zarr or any other Zarr URL, you can also set path to
the Zarr file in query parameters, for example:
https://datamedsci.github.io/ZoomingOnline/?data=http://localhost:8000/waveform.zarr
This project currently supports Zarr v2.
To add support for Zarr v3, migration to zarrita.js is required.
The project includes automated browser tests to verify application functionality.
📖 See testing.md for detailed information about testing tools and procedures.
MIT License — see LICENSE
The project was originally created as a part of the Large Scale Computing course at the AGH University in Kraków, Poland, by @ksew1, @irosikoni and @GosiaKk5.