Installation

pymccrgb is hosted on GitHub and PyPI. It requires LibLAS, Boost, and CMake and common Python packages like sklearn, skimage, and numpy.

LibLAS must be installed first:

sudo apt-get install liblas-c-dev

Then create a new virtual environment and install the package:

git clone https://github.com/rmsare/pymccrgb
cd pymccrgb
conda env create -f environment.yml
conda activate pymcc
pip install pymccrgb
py.test pymccrgb/tests

Conda installation

Conda forge package coming soon

You can install it with conda in a virtual environment.

conda env create -n pymcc
conda activate pymcc
conda install pymccrgb -c conda-forge

The package requires several C dependencies that may not be installed on your system. They are:

  • Boost
  • C++11 standard library
  • Cmake
  • LibLAS

The virtual environment will include all of these when the conda-forge package is installed.

pymccrgb is developed for Linux. The package can be installed on OS X, but it and Windows are not currently supported.

Installing LibLAS

LibLAS will be installed from conda-forge. If you need to install it manually, there are a couple of options.

If you are on a Debian-like Linux system, the LibLAS C API is available in the liblas-c-dev package. It can be installed with apt-get:

sudo apt-get install liblas-c-dev

Or from source

mkdir ~/liblas
cd ~/liblas
apt-get download liblas-c-dev  # Downloads the LibLAS .deb file
dpkg -x <deb-file> ~/liblas
export LD_LIBRARY_PATH=~/liblas/usr/include/:$LD_LIBRARY_PATH

On OS X, it is available as a Homebrew package.

brew install liblas

For more details, including how to compile LibLAS from source, see the LibLAS website.