Installing
Installing with pip (platform-independent)
Sardana can be installed using pip. The following command will automatically download and install the latest release of Sardana (see pip3 –help for options):
pip3 install sardana
Note that for sardana > 3.4.0, this will only install the minimal dependencies. To install all dependencies, use:
pip3 install "sardana[all]"
If you only want dependencies to run Spock and Sardana-Taurus Extension widgets:
pip3 install "sardana[spock,qt]"
You can test the installation by running:
python3 -c "import sardana; print(sardana.__version__)"
In sardana 3.6.0, the config tool was moved to its own repository. Install it with:
pip3 install sardana-config
Note: Installing sardana with pip3 might require building some dependencies like guiqwt or PyTango if no wheel is available for your platform. You will need a compiler. You could use Installing in a conda environment (platform-independent) to avoid this. If you decide to continue with pip3, please refer to PyTango’s installation guide. On Debian this should work to prepare the build environment:
apt-get install pkg-config libboost-python-dev libtango-dev
Linux (Debian-based)
Sardana is part of the official repositories of Debian (and Ubuntu and other Debian-based distros). You can install it and all its dependencies by doing (as root):
apt-get install python3-sardana
Note: python3-sardana package is available starting from the Debian 11
(Bullseye) release. For previous releases you can use python-sardana
(compatible with Python 2 only).
Installing in a conda environment (platform-independent)
In a conda environment (we recommend creating one specifically for sardana):
conda install -c conda-forge sardana
Since version 3.4.0, sardana is a metapackage that will install sardana and all optional
dependencies. For minimal requirements, install sardana-core instead.
See sardana-feedstock for the up-to-date list of packages.
Working from Git source directly (in develop mode)
If you intend to do changes to Sardana itself, or want to try the latest developments, it is convenient to work directly from the git source in “develop” (aka “editable”) mode, so that you do not need to re-install on each change.
Start by cloning the sardana repository:
git clone https://gitlab.com/sardana-org/sardana.git
cd sardana
In a conda or python virtual environment
You can use a python virtual environment or a conda environment with the needed dependencies:
# optional: if using a conda environment, pre-install dependencies with:
conda install --only-deps -c conda-forge sardana-core sardana-config taurus-qt
# install sardana in develop mode (inside the sardana repo)
pip3 install -e ".[all]" # <-- Note the -e !!
With pixi
Pixi is a package management tool for developers. Its goal is to provide developers with a clean and simple command-line interface to manage their project.
Pixi hides the complexity of installing dependencies and allows to run simple tasks whatever the operating system (Linux, Windows and macOS) using the conda ecosystem.
Install pixi if you haven’t already.
Pixi configuration is defined in the pixi.toml file at the root of the repository.
The required environment(s) will be created automatically when running a pixi command.
If you don’t have a Tango Database server running, run:
$ pixi run pydb
✨ Pixi task (pydb in pydb): PyDatabaseds 2
Ready to accept request
This will start a PyTango Database Device server, which uses SQLite to store data, on port 10000.
The previous command will automatically create an environment to install pytango-db and run PyDatabaseds 2.
It will create the file tango_database.db, where the data is saved.
In another terminal, run:
$ pixi run sar_demo
✨ Pixi task (sar_demo in default): sardanactl config load --write src/sardana/config/test/sar_demo.yaml
This will populate the database with the required Sardana instance and devices.
As long as you don’t delete the tango_database.db file, you don’t have to run this command again.
Once done, you can start Sardana:
$ pixi run Sardana
✨ Pixi task (sardana in default): Sardana demo1
Sardana is now running. In another window, you can start spock. The first time it will create an ipython profile
under ipython in the project directory. You’ll have to select to which door to connect:
$ pixi run spock
MainThread INFO 2025-04-23 16:59:27,145 TaurusRootLogger: Using PyQt5 (v5.15.10 with Qt 5.15.15 and Python 3.13.3)
Profile 'spockdoor' does not exist. Do you want to create one now ([y]/n)? y
Available Door devices from 127.0.0.1:10000 :
MainThread WARNING 2025-04-23 16:59:29,414 TaurusRootLogger: epics scheme not available: ModuleNotFoundError("No module named 'epics'")
Door_demo1_1 (a.k.a. Door/demo1/1) (running)
Door name from the list? Door_demo1_1
Storing ipython_config.py in /Users/johndoe/Dev/Sardana/sardana/.ipython/profile_spockdoor... [DONE]
Spock 3.5.3-alpha -- An interactive laboratory application.
help -> Spock's help system.
object? -> Details about 'object'. ?object also works, ?? prints more.
IPython profile: spockdoor
Connected to Door_demo1_1
Door_demo1_1 [1]:
With docker
The sardana-docker repository provides docker images mainly used for continuous integration. Those images can be used to run Sardana locally. Refer to the repository README.md for more information.
Dependencies
Sardana depends on PyTango, Taurus, lxml, itango and click. However some Sardana features require additional dependencies. For example: