Skip to content

Latest commit

 

History

History
87 lines (54 loc) · 2.5 KB

installation.rst

File metadata and controls

87 lines (54 loc) · 2.5 KB
Raw

Installation

Prospector itself is is pure python. To install a released version of just Prospector, use pip

python -m pip install astro-prospector

Then in Python

import prospect
print(prospect.__version__)

However, several other packages are required for the code to model and fit SEDs (see below.)

Development Version

To install the development version of Prospector and its dependencies to a conda environment, use the following procedure:

# change this if you want to install elsewhere;
# or, copy and run this script in the desired location
CODEDIR=$PWD
cd $CODEDIR

# Clone FSPS to get data files
git clone [email protected]:cconroy20/fsps
export SPS_HOME="$PWD/fsps"

# Create and activate environment (here named 'prospector')
git clone [email protected]:bd-j/prospector.git
cd prospector
conda env create -f environment.yml -n prospector
conda activate prospector
# Install latest development version of prospector
python -m pip uninstall astro-prospector
python -m pip install .

echo "Add 'export SPS_HOME=$SPS_HOME' to your .bashrc"

# To use prospector activate the conda environment
conda activate prospector

Requirements

Prospector works with python>=3.9, and requires numpy and SciPy

You will also need:

  • astropy
  • emcee and/or dynesty for inference (Please cite these packages in any publications)
  • sedpy (for filter projections)
  • HDF5 and h5py (If you have Enthought or Anaconda one or both of these may already be installed, or you can get HDF5 from homebrew or macports and h5py via pip)

For modeling galaxies you will need:

For parallel processing with emcee or dynesty (optional) you will need:

  • MPI (e.g. openMPI or mvapich2, available from homebrew, macports, or Anaconda) and mpi4py