You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 2 Next »

For Pandeia Engine v1.5.1

Requirements

The Pandeia Engine is a Python module, and requires both code and data files to operate.

Code

RequiredRecommended

python 2.7+

3.6.5

numpy 1.13.3+

1.15.4

scipy >1.4

1.1.0

astropy 2--3

2.0.9
photutils0.4
pysynphot0.9.12
setuptools40.8.0

Data

Required Data

In order to function, the Pandeia Engine requires one* set of reference data (containing instrument definitions):

James Webb Space Telescope: https://stsci.app.box.com/v/pandeia-refdata-v1p5p1

Nancy Grace Roman Space Telescope: https://stsci.box.com/v/pandeia-refdata-v1p5p1-wfirst

The location of the top of the data file tree (the folder containing PSF_VERSION) must be set as the $pandeia_refdata environment variable.

*Both telescope datasets may be untarred to the same location to produce an engine that can work with either telescope; any files the two packages have in common may be overwritten.

Optional Data

An additional dataset can be used to provide more spectroscopic functionality (phoenix spectral models, normalization by HST bandpasses, normalizing in among others).

The files will produce a directory tree of grp/hst/cdbs; the Pandeia Engine expects the environment variable $PYSYN_CDBS to point to the cdbs directory.

Installation

We recommend using the Anaconda python distribution. Follow its installation instructions to set up a new environment if you do not already have one.

Once installed, from a terminal with access to the conda environment, type "conda config --add channels http://ssb.stsci.edu/astroconda" to install the Astroconda channel.

Create an AstroConda environment with the STScI package installed. For example, to create an environment named "pandeia" working under python 3.6, type

conda create -n pandeia stsci numpy scipy=1.1 astropy=2 photutils pysynphot setuptools python=3.6

from a terminal with access to the conda environment.

Then, install the engine itself: `pip install pandeia.engine==1.5.1`


It should also be possible to install the Pandeia Engine and all its dependencies by simply running `pip install pandeia.engine==1.5.1` but this is not recommended and users are likely to run into more issues if it is installed this way.

Verify Installation

Open up a terminal with access to the conda installation, and type

python -c "import pandeia.engine; pandeia.engine.pandeia_version()"

If properly installed and configured, it should show the refdata version and synphot data directory.

Helper Script

The following script assumes you are operating in a terminal where a conda installation has been activated, that you've installed the pandeia refdata in ${HOME}/data/pandeia, and the cdbs folder from the additional dataset moved and renamed to ${HOME}/data/pysynphot

Installation script
conda config --add channels http://ssb.stsci.edu/astroconda
conda create -n pandeia stsci numpy scipy=1.1 astropy=2 python=3.6
mkdir -p ~/anaconda/envs/pandeia/etc/conda/activate.d
cat >>~/anaconda/envs/pandeia/etc/conda/activate.d/env_vars.sh <<EOF
#!/bin/sh
export PYSYN_CDBS=${HOME}/data/pysynphot
export pandeia_refdata=${HOME}/data/pandeia
EOF
mkdir -p ~/anaconda/envs/pandeia/etc/conda/deactivate.d
cat >>~/anaconda/envs/pandeia/etc/conda/deactivate.d/env_vars.sh <<EOF
#!/bin/sh
unset PYSYN_CDBS
unset pandeia_refdata
EOF
conda activate pandeia
pip install pandeia.engine==1.5.1

This creates a pandeia environment in an already-existing conda installation, and sets up two scripts to automatically set and unset the required environment variables every time you activate and deactivate the pandeia environment.


  • No labels