Currently, to install the Python binding of this package, one has to use pip ( if they don't want to install from source). However, the installation requires compiling, so it requires certain versions of cmake and gcc etc. They have to be installed before running pip install, or the pip installation could fail.
To manage these build-time dependences, another option to provide is Anaconda. The version requirements on compilers etc can be expressed in the anaconda environment, so the users do not have to face the potential of installation failure due to version requirements on compilers etc.
Currently, to install the Python binding of this package, one has to use
pip( if they don't want to install from source). However, the installation requires compiling, so it requires certain versions of cmake and gcc etc. They have to be installed before runningpip install, or the pip installation could fail.To manage these build-time dependences, another option to provide is Anaconda. The version requirements on compilers etc can be expressed in the anaconda environment, so the users do not have to face the potential of installation failure due to version requirements on compilers etc.