Installation

TimeFeatures is an Orange3 add-on, so it requires a working Orange3 installation in the same Python environment.

From PyPI

pip install TimeFeatures

This pulls in every runtime dependency, including the precompiled psycopg2-binary wheel so the Save to DB widget works on macOS, Linux and Windows without a C toolchain.

From source

git clone https://github.com/alervgr/Orange-TimeFeatures.git
cd Orange-TimeFeatures
pip install -e .

The -e flag installs in editable mode so changes to the source tree are picked up on the next Orange restart without reinstalling.

Inside Orange

TimeFeatures also installs through Orange’s GUI:

1. Open Orange.

2. Go to Options → Add-ons….

3. Tick TimeFeatures in the list and confirm.

4. Restart Orange. The new widgets appear under the Time-Features toolbox section.

Anaconda

The same pip install TimeFeatures command works inside a Conda environment as long as Orange3 is installed there.

Dependencies

TimeFeatures declares the following runtime dependencies (see setup.py):

Package	Minimum version	Used by
numpy	1.22.4	All widgets.
scipy	1.7.3	Sparse adjacency matrices in the dependency graph.
PyQt5	5.15.6	Widget UI.
AnyQt	0.2.0	Qt abstraction used by Orange.
SQLAlchemy	1.4.0	Dialect-agnostic SQL toolkit used by Save to DB.
psycopg2-binary	2.9.9	PostgreSQL driver for Save to DB.
PyMySQL	1.0.0	MySQL driver for Save to DB.
Orange3-Network	1.8.0	Network type produced by Variable Dependency Graph.

Orange3 itself is intentionally left out of the requirements list — the host application provides it.

Running tests

The test suite uses unittest and lives under timefeatures/widgets/tests/. From a checkout:

pip install -e .
python -m unittest discover -v timefeatures/widgets/tests

The widget-level tests inherit from Orange.widgets.tests.base.WidgetTest and therefore require Qt; a virtual display (Xvfb) is enough on a headless CI worker.