Convert code blocks to basic Markdown style. authored by Alan De Smet's avatar Alan De Smet
......@@ -16,37 +16,29 @@ You will need to have the python 2.7 version of Anaconda installed for these ins
On SSEC servers you use this command to load a pre-installed version of Anaconda:
~~~~
module load anaconda27
~~~~
Once you have Anaconda installed, run the following commands on the command line:
~~~~
conda create -n glance python=2.7 anaconda
source activate glance
conda install basemap netcdf4 h5py pyqt=4.11.4 mock scipy pillow
conda install -c http://sips.ssec.wisc.edu/conda python-hdf4
# on Mac OSX, you will need to use "conda install -c NSIDC python-hdf4" instead
python -m easy_install -U -vi http://larch.ssec.wisc.edu/eggs/repos uwglance
~~~~
This will create an Anaconda environment called "glance" that can be used to run Glance at the command line. When you open a new terminal session with your computer to use glance you will need to activate this environment using the command:
~~~~
source activate glance
~~~~
If you need a developer installation of Glance, you will follow most of the same instructions as above, but you will also need to be able to use git on the command line.
Replace the final "python -m easy_install ..." command in the installation instructions above with:
~~~~
(cd to the directory where you want the Glance code installed)
git clone git@gitlab.ssec.wisc.edu:evas/UW-Glance.git
cd ./UW-Glance/pyglance
python setup.py develop
~~~~
Installing with Minimal Dependencies Using Pip
----------------------------------------------
......@@ -59,16 +51,13 @@ You will need development libraries for libpng, libfreetype, libgeos, and pyqt4.
Ubuntu 16.04 installation of the dependencies:
~~~~
sudo apt-get install libpng-dev libfreetype6-dev pyqt4-dev-tools libgeos-dev virtualenv python-pip
# Ubuntu has problem with the libgeos library's name
cd /usr/lib/x86_64-linux-gnu
sudo ln -s libgeos-3.5.0.so libgeos.so
~~~~
Glance installation:
~~~~
# Use system-site-packages to use distribution's pyqt4
virtualenv --system-site-packages glance
cd glance
......@@ -77,20 +66,16 @@ pip install scipy matplotlib pillow
# Not available via PyPI
pip install https://github.com/matplotlib/basemap/archive/v1.0.7rel.tar.gz
pip install http://larch.ssec.wisc.edu/eggs/repos/uwglance/uwglance-0.3.2.tar.gz
~~~~
If you want to do development on Glance, skip the "pip install ...uwglance..." step and instead:
~~~~
git clone https://gitlab.ssec.wisc.edu/evas/UW-Glance.git
cd uwglance/pyglance
python setup.py develop
~~~~
Minimal Anaconda install, no GUI
--------------------------------
~~~~
mkdir glance
cd glance
wget https://repo.continuum.io/miniconda/Miniconda2-latest-Linux-x86_64.sh
......@@ -100,7 +85,6 @@ conda install -y numpy scipy matplotlib pillow geos basemap mako markupsafe netc
git clone --depth=1 https://gitlab.ssec.wisc.edu/evas/UW-Glance.git
cd UW-Glance/pyglance/
python setup.py develop
~~~~
The resulting executable is in glance/anaconda/bin/glance
......@@ -113,22 +97,20 @@ Installing Using HHG
Please don't try to do this unless you already know what HHG is.
Prerequisites:
~~~~
# make sure PYTHONPATH is not set in environment (remove from .bashrc / .bash_profile / etc)
unset PYTHONPATH
# check that SSEC SB3-hhg 2014Q1 is installed, should be r3 or newer.
rpm -qa |grep shellb3-2014q1
# make sure $HOME/opt exists
mkdir -p $HOME/opt
~~~~
Installation uses 12M of disk space:
~~~~
# make a virtualenv to place glance in, using SB3-hhg as base runtime
/opt/ShellB3/2014Q1/bin/virtualenv --system-site-packages $HOME/opt/glance.env
# add latest glance to that virtualenv
$HOME/opt/glance.env/bin/python -m easy_install -U -vi http://larch.ssec.wisc.edu/eggs/repos uwglance
~~~~
Usage (there are 3 options here, pick 1) :
1. softlink "$HOME/opt/glance.env/bin/glance" into your PATH
......
......