This section will provide you with information to establish a python environment in order to develop qudi.

Do not use the following installtion quide, if you want to install Qudi for normal use. Check the Installation page to see how to do that.

Installation page with full install instructions

How to install packages in python

Additional packages can be installed either via conda or via pip from command line (you might need admin rights).

Try at first to use the conda package manager (which comes together with the Anaconda distribution):

conda install <modulename>

That is the most elegant way, since conda will handle all cross dependencies with other packages, will download all needed packages for that module and will try to resolve appearing issues. If <modulename> is not in the repository list of Anaconda, you can try to install the module via pip, the internal package manager for python:

pip install <modulename>

Note that pip will not look for cross dependencies but just simply install and expand the desired package.

Pre-requisition

It is assumed, that you have already selected the a conda environment, for which the package installation will be performed. All the listed commands are executed in the terminal or cmd (for windows) with administrator privileges (!).

For more information concerning the conda environment see: https://conda.io/docs/user-guide/tasks/manage-environments.html

By default, if anaconda is installed on your machine, the installation will happen in the environment "base". In order to create a separate environment with the name "qudi" perform

conda create -n qudi python=3.6

You can choose also a different name than 'qudi' for the environment (and the desired python version), but you have to replace then 'qudi' by your custom name in the following process.

Conda package installation

You can install each package separately

conda install -n qudi cycler
conda install -n qudi cython
conda install -n qudi ipython
conda install -n qudi jupyter
conda install -n qudi lxml
conda install -n qudi matplotlib
conda install -n qudi numpy
conda install -n qudi pillow

or all bundled all in one command:

conda install -n qudi cycler cython ipython jupyter lxml matplotlib numpy pillow

Those are the packages, which are precompiled and/or distributed from the Anaconda package index via the conda package tool. Note that all related packages (e.g. pyqt and qt bindings,...) will be automatically installed during this procedure.

Pip package installation

Pip will install now specific packages, which are required for qudi. Before you perform the installation with the internal package manager, you have to activate the "qudi" environment. In the CMD perform

activate qudi

For linux users, the command would be

source activate qudi

Now you can perform either an installation of each package

pip install asteval
pip install fysom
pip install gitpython    
pip install lmfit
pip install pydaqmx
pip install pyflowgraph-qo
pip install pyqtgraph-qo 
pip install pyvisa
pip install rpyc
pip install ruamel.yaml
pip install serial
pip install typing

or install all in one run

pip install asteval fysom gitpython lmfit pydaqmx pyflowgraph-qo pyqtgraph-qo pyvisa rpyc ruamel.yaml serial typing

This should install the correct environment for qudi.

Extract current environment to file

If you want to extract the current conda environment 'qudi' to file, just perform

conda-env export -n qudi > my-env.yml

Note that in the file 'my-env.yml', there will be at the bottom the section 'prefix', which identifies the path to the conda environment. Make sure to delete the section 'prefix', if you want to use this file to install the 'qudi' environment elsewhere, to prevent confusions about the location of the conda environment. Then, in the installation procedure, the default location of the conda environment will be taken, which can be different on every machine.

Install conda environment from file

If you have a conda environment file 'my-env.yml' present, you can use that to create manually the conda environment. All the required information should be written in the file 'my-env.yml'.

conda env create -f my-env.yml

Remove environment

If you want to remove the environment with the name 'qudi', just perform

conda-env remove --yes --name qudi