Skip to content

Latest commit

 

History

History
136 lines (110 loc) · 3.85 KB

Setup.md

File metadata and controls

136 lines (110 loc) · 3.85 KB
title layout
Setup
default

Setup

Please make sure to install everything before the start of your workshop and run the tests. Participants must bring and use their own laptops to insure the proper setup of tools for an efficient workflow once you leave the workshop.

Please note that if you decide not to follow the below instructions and use local installs of your software it is not guaranteed that the summer school team can assist in debugging any issues that may arise.

Conda

Conda is an open source package management system and environment management system for installing multiple versions of software packages and their dependencies and switching easily between them. It works on Linux, OS X and Windows, and was created for Python programs but can package and distribute any software.

To install conda please run the following commands from a bash shell. If you already have conda installed please run conda update conda and move on to the next step.

Mac instructions:

curl https://repo.continuum.io/miniconda/Miniconda3-latest-MacOSX-x86_64.sh -O
bash Miniconda3-latest-MacOSX-x86_64.sh -b -p $HOME/miniconda
echo PATH="\$HOME/miniconda/bin:\$PATH" >> ~/.bash_profile

Linux instructions (or Windows Ubuntu shell):

wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh -O miniconda.sh
bash miniconda.sh -b -p $HOME/miniconda
echo PATH="\$HOME/miniconda/bin:\$PATH" >> ~/.bashrc

The Software Summer School environment

To make a great deal of compilation and installation simpler we will create a conda environment. This environment isolates the software summer school stack from the rest of the dependencies on your laptop and avoid compilation of code.

Mac conda environment

First run:

sysctl machdep | grep features | grep -o AVX2

If the above returns "AVX2" please use the following lines:

conda create -n sss python=3.5 psi4 numpy lawrap gnu cmake jupyter scipy numexpr mkl-include gcc-5-mp -c intel -c psi4/label/dev -c psi4

otherwise use the below:

conda create -n sss python=3.5 psi4 numpy lawrap gnu cmake jupyter scipy numexpr mkl-include gcc-5-mp sse41 -c intel -c psi4/label/dev -c psi4

Linux conda environment

conda create -n sss python=3.5 psi4 numpy lawrap cmake jupyter scipy numexpr mkl-include gcc-5-mp -c intel -c psi4/label/dev -c psi4

Testing

Please ensure that the installation was successful by trying out both of the tests below.

Test the Psi4 installation

Test the installation:

source activate sss
python -c 'import psi4; psi4.test()'

As long as there are no tests that end with a Failed! the installation was successful. Note xfailed is fine (an expected fail).

Test the compilers

First copy the following C++ code to a file named thread_test.cpp:

#include <iostream>
#include <omp.h>
int main() {
    #pragma omp parallel num_threads(4)
    {
        #pragma omp critical
        std::cout << "tid = " << omp_get_thread_num() << std::endl;
        std::cout << "Hello World!" << std::endl;
    }
}

Mac compile line:

In order for the mac compilers to work correctly xcode command line tools need to be installed. Before compiling the below please run xcode-select --install.

g++ -o thread_test -fopenmp thread_test.cpp -Wl,-rpath,${CONDA_PREFIX}/lib/

Linux compile line:

g++ -o thread_test -fopenmp thread_test.cpp

You can then run this compiled executable and you should see the following (number ordering may change):

user:~/tests ./thread_test
tid = 2
Hello World!
tid = 0
Hello World!
tid = 3
Hello World!
tid = 1
Hello World!

Acknowledgements

The above work could not have been possible without the help from many open-source projects. In particular, most of the above ecosystem is based of (and supported by) the excellent work of Lori Burns at the Georgia Institute of Technology.