huzizhan / ipycles

A python based infrastructure for cloud large eddy simulation and stable water isotope

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

iPyCLES project is a Large Eddy Simulation model (PyCLES) coupled with stable water isotope tracer components, developed by the Jonathon S. Wright group in the Department of Earth System Science, Tsinghua University. Now the isotopic components adopted under the branch isotopetracer with the model version is iPyCLES v1.0. The structure of iPyCLES v1.0 is like the following schematic diagram:

Installation of ipycles in Linux (tested with ubuntu, Debian and Centos):

Important system environments needed to be installed:

$ sudo apt-get install gcc gfortran-8 csh for Debian based distributions or $ sudo yum install gcc csh for Centos based distributions: like Fedora, Alamlinux, etc.

Here the gfortran version should be lower than 9 (the lastest version until May, 2021), or some files can't be compiled.

We recommend using conda as the package management system and environment management system for python environment settings. Miniconda can be downloaded using wget:

$ wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh

then install by:

$ bash Miniconda3-latest-Linux-x86_64.sh

Becareful with the python version, which can't be higher then 3.8, because cython doesn't support python 3.9 until now (May, 2021).

Install packages needed to compile ipycles:

$ conda install numpy scipy netcdf4 mpi4py matplotlib cython gcc_linux-64

Complie ipycles by doing following steps:

  1. $ cd ipycles

  2. $ python generate_paramters.py

  3. $ CC=mpicc python setup.py build_ext --inplace

Besides, if there is no openmpi environemnt, you can also try to install 'mpich-mpicc' using conda like: $ conda install -c conda-forge mpich-mpicc and then run though the complie processes above

More details about installation in defferent platforms and erros can be found : install.rst

Run test cases

Run test cases of ipycles follows: running.rst

More information about PyCLES can be found in pycles

About

A python based infrastructure for cloud large eddy simulation and stable water isotope

License:GNU General Public License v3.0


Languages

Language:Fortran 88.6%Language:Python 7.5%Language:C 2.9%Language:Cython 1.0%Language:Shell 0.0%