Tiled Qt

A modern desktop GUI client for the Tiled scientific data service, built with PyQt5 and Matplotlib.

Features

• Connect to any Tiled server (public demo, local, or institutional).
• Browse the data tree with lazy-loading and efficient pagination.
• Metadata Viewer: Inspect nested metadata in a hierarchical tree view.
• Array Visualization: Plot 1D arrays, 2D images, and 3D slices with Matplotlib.
• Table Viewer: Browse DataFrames with a scrollable grid view, optimized for large datasets.
• Advanced Search: Query metadata using Key-Value pairs, Full Text, and Regex.
• Export: Save data and containers to HDF5, CSV, Excel, Parquet, PNG, TIFF, NumPy, and more.
• Dynamic Format Discovery: Export options are automatically filtered based on what the Tiled server supports for each specific item.
• On-Demand Loading: Fetch data only when needed, with support for remote slicing.
• Navigation: Full history support (Back/Forward) and breadcrumb navigation.
• Security: Safe slice parsing (no eval) and non-interactive authentication.

Installation

Using Pixi (Recommended)

If you have Pixi installed, you can run the application directly:

pixi run tiled-qt

To run in debug mode:

pixi run tiled-qt debug

Using Pip

You can install the package in editable mode from the source:

pip install -e .

This will provide the tiled-qt command in your environment.

Using Rattler (Conda/Conda-forge)

The project includes a recipe.yaml for building a conda package using rattler-build:

rattler-build build --recipe recipe.yaml --target-platform linux-64

Usage

After installation, simply run:

tiled-qt

Quick Start

1. Connect: Click ⚡ Connect or press Ctrl+K.
2. Select Server: Use the NSLS-II Demo preset or enter your server URL.
3. Browse: Navigate the tree on the left. Containers load children lazily.
4. Inspect: Click an item to view its metadata and basic information.
5. Visualize: Click Load Data or Load Table to fetch and display the data.
6. Slicing: For large arrays, enter a slice (e.g., 0, :, :) before loading.
7. Export: Right-click on a tree node or use the Export button in viewers. The available formats (HDF5, CSV, PNG, etc.) are dynamically discovered from the server.
8. Search: Use the Search tab to find specific data by metadata.

Development

The project is structured as a modular Python package:

• src/tiled_qt/app.py: Main application window and logic.
• src/tiled_qt/widgets/: Specialized UI components (Viewers, Search, etc.).
• src/tiled_qt/styles.py: UI styling and themes.
• src/tiled_qt/utils.py: Logging and helper functions.

Running with a package manager

# Using Pixi
pixi run tiled-qt

# Using standard Python
python3 -m tiled_qt

Demo Server

You can test the application using the NSLS-II public demo server: https://tiled-demo.nsls2.bnl.gov (No login required).

License

This project is licensed under the BSD 3-Clause License.