Installation
=============================

System requirements
-------------------------
- Python (version >= 3.6).
- (For linux only) X11 system. Make sure other GUI programs (such as xeyes) works properly. For Windows users using Windows Subsystem for Linux (WSL), it is recommended to use xvcsrv (https://sourceforge.net/projects/vcxsrv/) as X11 client. 

Before installation
--------------------------
It is encouraged to use virtual environment to avoid conflict. You can skip this section if you want to install lys directly.

1. Open shell (we assume bash in linux). Go to home directory (or arbitrary directory you want to create virutual environment).

2. Create .venv directory and create "lys" (or arbitrary name you want) environment::

    mkdir .venv
    cd .venv
    python -m venv lys

3. Activate the environment::

    source lys/bin/activate

Installation from pip (recommended)
----------------------------------------------------

1. Update pip::

    pip install --upgrade pip

2. Install lys by pip::

    pip install lys-python

3. Start lys by the command below. Note that the current directory of the system is used as the working directory of lys::

    python -m lys

4. If you want to enable parallel computing, start *lys* with -n option::

    python -m lys -n [number of cores]

Installation from source (not recommended)
--------------------------------------------------------

If you want to install `lys` from source, follow the instructions below.
In particular, this is recommended if you want to use :doc:`tutorials/test`.

1. Update pip::

    pip install --upgrade pip

2. Clone lys. If you do not have git, you can download the source code from GitHub (https://github.com/lys-devel/lys)::

    git clone git@github.com:lys-devel/lys.git

3. Install lys by pip. If you want to install lys in development mode, add `-e` option after `pip install`::

    cd lys
    pip install .

4. (Optional) If you want to check installation, go to :doc:`tutorials/test`.

5. Start lys by the command below. Note that the current directory of the system is used as the working directory of lys::

    python -m lys

6. If you want to enable parallel computing, start *lys* with -n option::

    python -m lys -n [number of cores]

Library version
-------------------------

We confirmed that *lys* works well under the environment below. If *lys* does not work, try the library versions below.

Python 3.6.8

- numpy 1.19.5
- scipy 1.5.4
- opencv-python-headless 4.5.5.64
- dask 2021.3.0
- dask-image 2021.12.0
- matplotlib 3.3.4
- pyqtgraph 0.11.1
- PyQt5 5.15.6
- qtpy 2.0.1
- autopep8 1.6.0


Python 3.11.5

- numpy 1.26.0
- scipy 1.11.3
- opencv-python-headless 4.8.1.78
- dask 2023.9.3
- dask-image 2023.8.1
- matplotlib 3.8.0
- pyqtgraph 0.13.3
- PyQt5 5.15.9
- qtpy 2.4.0
- autopep8 2.0.4