astroNN APOGEE VAC codes
This repository contains codes required to generate astroNN APOGEE VAC
- astroNN DR16 VAC: https://www.sdss.org/dr16/data_access/value-added-catalogs/?vac_id=the-astronn-catalog-of-abundances,-distances,-and-ages-for-apogee-dr16-stars
- astroNN DR1 7VAC: To be released publicly
Requirement
python>=3.6
, numpy
, scipy
, astropy
, astroquery
, galpy
, astroNN
, pyyaml
, statsmodels
, h5py
, gaia_tools
zero_point
are required to correct Gaia zeropoint: https://gitlab.com/icc-ub/public/gaiadr3_zeropoint
Scripts
Please review the configurations stored in the following following files before running any script.
- Global configuration for the scripts
- Gaia login if you have any
- Contains a few useful utility functions
- A simple script to retrieve all ASPCAP normalized spectra and save to a single file
Please run the scripts in ASCENDING ORDER
- This script continuum normalizing the original data set that the models trained on (aka dr14)
- This script continuum normalizing all the APOGEE spectra and save them to a single fits
- This script cross matching APOGEE-Gaia to get all columns and save them to a single fits
- This script uses neural network models to get abundances, distance, ages and save them to fits
- This script uses galpy to generate orbital parameters and save them to fits
- This script uses all the fits files we have generated and compile VAC
- This script make some summary plots for the VAC
External Data
APOKASC_cat_v6.6.1.fits.zip
: https://trac.sdss.org/attachment/wiki/APOGEE2/APOKASC/Catalog/APOKASC_cat_v6.6.1.fits.zipkepler_low_metallicity_with_samples.fits
: Internal use only
DR16 VAC
The folder astroNN_0617_run001
, astroNN_constant_model_reduced
, APOKASC2_BCNN_age_only_corrections0.1
are original neural network models used in DR16 VAC
DR17 VAC
We have retrained our models for DR17 VAC, the training scripts and sanity checks notebook are under the folder dr17-VAC-notebooks
The folder astroNN_0512_run002
, astroNN_gaia_dr17_model_3
, APOKASC2_BCNN_age_combined_dr17_4
are new neural network models used in DR17 VAC
lco_apo_median_dr17sync.npy
is the clculated median difference between apo25m and lco25m
Major Authors
- Henry Leung - henryskyContact Henry: henrysky.leung [at] utoronto.ca
- Jo Bovy - jobovy
- Ted Mackereth - jmackereth