Linux

Python - general

Install a Python 3 environment, e.g. Anaconda or Virtualenv, and make sure the following Python packages are installed:

  • numpy, matplotlib, pandas, sphinx

For the case of missing packages you can either use conda (provided the Anaconda stack was installed)

conda install numpy matplotlib scipy pandas sphinx

or pip

pip install numpy matplotlib scipy pandas sphinx

How to setup Anaconda

Anaconda can be downloaded here: https://www.anaconda.com/download/ An installation guide for Linux (and for other platforms) can be found here: https://docs.anaconda.com/anaconda/install/linux

How to setup Virtualenv

Virtualenv is a tool to create isolated Python environments. An installation guide and all virtualenv-support can be found here: https://virtualenv.pypa.io/en/stable/installation/

Creat a virtual environment

mkvirtualenv --python /usr/bin/python3 NAME

with NAME as the name of your virtual environment. For e.g. CRTomo tool you can set NAME to crtomo or for EDF to edf. Open your virtual environment with

workon *NAME*

Setup a working directory

Create the directory: /home/USERNAME/ccd-tools and subdirectories for the different git repositiries.

How to use the Git repository

An installation guide for Git can be found here: https://git-scm.com/book/en/v2/Getting-Started-Installing-Git Install git and get yourself familiar with the following git commands:

  • git clone

  • git pull / git push

  • git add

  • git commit

  • git branch

  • git checkout

  • git merge

Simply clone the following repositories:

git clone https://github.com/m-weigand/sip_models.git
git clone https://github.com/m-weigand/geccoinv.git
git clone https://github.com/m-weigand/ccd_tools.git

into the corresponding folder /home/USERNAME/ccd-tools.

Getting Started: CCD-tools with Anaconda

Getting Started: CCD-tools with Virtualenv

Python and Virtualenv

1. Install Python and Virtualenv as described above. 3. Create a virtual environment (named ccd-tools), using the following commands:

mkvirtualenv --python /usr/bin/python3 ccd-tools
pip install --upgrade pip

If the virtual environment already exists, you can enter it, using the following command:

workon ccd-tools

A) Installation of ccd-tools and required packages via pip install

Enter your virtual environment and install ccd-tools including sip-models and geccoinv, using the following commands:

workon ccd-tools
pip install ccd_tools

For the installation via pip install is no local copy of the git-repositories needed.

B) Installation of ccd-tools and required packages via local source

  1. Install Git as described above.

  2. Create a working directory and clone the Git repositories into the corresponding folder.

  3. Enter your virtual environment, using the following commands:

    workon ccd-tools

  4. To install geccoinv, go to the corresponding directory …/ccd-tools/geccoinv and use the following commands:

    pip install -U pip
pip install -r requirements.txt
python setup.py install

  5. To install sip-models, go to the corresponding directory …/ccd-tools/sip_models and use the following commands:

    pip install -r requirements.txt
python setup.py install

  6. To install ccd-tools, go to the corresponding directory …/ccd-tools/ccd_tools and use the following commands:

    pip install -r requirements.txt
python setup.py install

Using a user-defined directory

The package can also be installed to a user-defined directory:

export PYTHONUSERBASE=$HOME/inst/pip_installs
export PYTHONPATH=$HOME/inst/pip_installs/lib/python3.4/\
    site-packages/:$PYTHONPATH
python setup.py install --user
export PATH=$HOME/inst/pip_installs/bin:$PAT

Getting Started: ccd-tools with Jupyter Notebooks and Virtualenv

Installing Jupyter Notebook

  1. Install Python and Virtualenv, create a virtual environment and install ccd-tools as described above (A or B, both possible)

  2. Enter your virtual environment and install Jupyter Notebooks, using the following commands:

    pip3 install --upgrade pip
pip3 install jupyter

  3. Install and enable the Jupyter Widget JavaScript library, using the following commands:

    pip install ipywidgets
jupyter nbextension enable --py widgetsnbextension --sys-prefix

Starting the Notebook Server

Navigate to the Notebook example directory …/ccd-tools/ccd_tools/Examples/notebooks and open Jupyter Notebooks:

jupyter notebook

This will print some information about the notebook server in your terminal, including the URL of the web application (by default, http://localhost:8888) It will then open your default web browser to this URL. When the notebook opens in your browser, you will see the Notebook Dashboard, which will show a list of the notebooks, files, and subdirectories in the directory where the notebook server was started:

…/ccd-tools/ccd_tools/Examples/notebooks

Using the Notebooks

Supplementary installation instructions

Installation for developers:

python seutp.py develop --user

To build the documentation:

cd docs/doc
make html

setuptools developer guide: https://pythonhosted.org/setuptools/setuptools.html

Old installation guide

The setuptools distribution tools to manage the installation procedure:

python setup.py install

should suffice to install the libraries and scripts.

python setup.py build
python setup.py install --prefix=$HOME/inst/dd

export PYTHONUSERBASE=$HOME/inst/pip_installs
export PYTHONPATH=$HOME/inst/pip_installs/lib/python2.7/\
    site-packages/:$PYTHONPATH
python setup.py install --user
export PATH=$HOME/inst/pip_installs/bin:$PATH
python seutp.py develop --user

To build the documentation, execute

cd docs/doc
python setup.py sphinx_build

For certain versions of numpy (Debian Wheezy), there exist problems with libopenblas (CPU goes to 100% and the program freezes). These problems are related to multithreading issues in Python

Workarounds:

  • Use only one thread in openblas:

    OPENBLAS_NUM_THREADS=1 ccd_single [...]

  • switch to atlas/libblas:

    update-alternatives --config libblas.so
update-alternatives --config libblas.so.3