Installation & Setup 

FastSpecFit can be set up in two ways:

At NERSC — for DESI collaborators, using the shared software stack.
On a laptop — for interactive work or running on smaller datasets.

NERSC Installation 

To run FastSpecFit at NERSC, first request an interactive CPU node:

salloc -N 1 -C cpu -A desi -t 01:00:00 --qos interactive

Then, load the package on top of the standard DESI software stack:

source /dvs_ro/cfs/cdirs/desi/software/desi_environment.sh main
module load fastspecfit/main

To load specific tagged versions, replace main with the desired release, for example:

source /dvs_ro/cfs/cdirs/desi/software/desi_environment.sh 26.3
module load fastspecfit/3.4.3

To work with a local git clone of FastSpecFit at NERSC, first load the DESI software stack as above (without loading the fastspecfit module), then run the following once after cloning or after each git pull:

pip install --no-deps -e /path/to/fastspecfit

This registers the CLI entry points (fastspec, fastphot, etc.) in the active conda environment and generates the _version.py file that records the correct package version in output file headers. The --no-deps flag skips reinstalling dependencies already provided by the DESI software stack.

Alternatively, in some cases it may be necessary to bypass the pip install step and manually add FastSpecFit to your PYTHONPATH and PATH via:

export PYTHONPATH=/path/to/fastspecfit/py:${PYTHONPATH}
export PATH=/path/to/fastspecfit/bin:${PATH}

JupyterHub 

To use FastSpecFit within NERSC’s JupyterHub, first install the appropriate Jupyter kernel (one time). Then, to access the main branch, do:

mkdir -p ${HOME}/.local/share/jupyter/kernels/fastspecfit-main
wget -O ${HOME}/.local/share/jupyter/kernels/fastspecfit-main/kernel.json \
  https://raw.githubusercontent.com/desihub/fastspecfit/main/etc/jupyter-kernel.json

Or, to use the latest tagged release, do:

mkdir -p ${HOME}/.local/share/jupyter/kernels/fastspecfit-latest
wget -O ${HOME}/.local/share/jupyter/kernels/fastspecfit-latest/kernel.json \
  https://raw.githubusercontent.com/desihub/fastspecfit/main/etc/jupyter-kernel-latest.json

Then select the FastSpecFit Main or FastSpecFit Latest kernel from the JupyterHub launcher.

Laptop Installation 

Create a Virtual Environment 

We recommend micromamba for creating isolated Python environments, though Miniforge and Anaconda work equally well.

micromamba create -n fastspec python
micromamba activate fastspec

Install FastSpecFit 

FastSpecFit is available on PyPI and installs with a single command:

pip install fastspecfit

Data Dependencies 

Running FastSpecFit on DESI data requires four additional data products, each pointed to by an environment variable.

DESI_SPECTRO_REDUX: Root directory of the DESI spectroscopic reduction outputs. The DESI products are organized according to the DESI data model; the DESI data organization page describes the full on-disk layout. The DESI Early Data Release (EDR) and Data Release 1 (DR1) are already publicly available, with Data Release 2 (DR2) expected to be public in early 2027.
DUST_DIR: Location of the Schlegel, Finkbeiner, & Davis (1998) dust reddening maps, used to correct spectra and photometry for Milky Way dust extinction. The maps can be downloaded from the NERSC public portal (see commands below).
FPHOTO_DIR: Location of the DESI Legacy Imaging Surveys DR9 broadband photometry, which provides optical and near-infrared fluxes used for joint spectrophotometric SED fitting. See the DR9 description for an overview of the imaging data and data products.
FTEMPLATES_DIR: Location of the stellar population synthesis template files used by FastSpecFit to model the stellar continuum. Templates are distributed via the DESI public data portal and can be downloaded with the command below.

Set these variables and download the required external files:

export DESI_SPECTRO_REDUX=/path/to/spectro/redux
export DUST_DIR=/path/to/dustmaps
export FPHOTO_DIR=/path/to/dr9/data
export FTEMPLATES_DIR=/path/to/templates/fastspecfit

wget -r -np -nH --cut-dirs 5 -A fits -P $DUST_DIR \
  https://portal.nersc.gov/project/cosmo/data/dust/v0_1/maps
wget -O $FTEMPLATES_DIR/ftemplates-chabrier-2.0.0.fits \
  https://data.desi.lbl.gov/public/external/templates/fastspecfit/2.0.0/ftemplates-chabrier-2.0.0.fits

Building the Documentation 

To build the documentation locally, install the optional documentation dependencies and run Sphinx:

pip install "fastspecfit[doc]"
sphinx-build -W --keep-going -b html doc doc/_build/html