CAFExtractor (hereafter cafex) is the observatory pipeline of the upgraded CAFE instrument (called CAFE2, see the instrument webpage here). This pipeline is meant to be used by the observatory every night. The purpose is that CAFE2 users can have their data reduced by the morning after the observations. The pipeline and the new upgrade of the instrument is presented in Lillo-Box et al. (2019, MNRAS). CAFExtractor is partly based on the CERES pipeline (Brahm et al., 2017).
Version: v0.8
Before starting, make sure you have the following python packages installed:
python 2.7, numpy, scipy, pyfits, astropy, termcolor, ntpath, argparse
And:
apt install python-pip python-pycurl libgsl-dev
Some packages require specific versions. In particular:
pip install --user pandas==0.19.2 scipy==1.17.1 statsmodels==0.6.1
Just clone this directory to a local path like:
mkdir /path_to_folder/cafextractor
cd /path_to_folder/cafextractor
git clone https://github.com/jlillo/cafextractor.gitThen make sure you tell you computer where the pipeline is:
setenv PYTHONPATH ${PYTHONPATH}:/path_to_folder/cafextractor/srccafeX does not need anything else to work but some modules from the CERES pipeline require that you compile some fortran routines. To that end, please follow the instructions properly described in the Installation section of the CERES pipeline here. Basically, you have to foolow these steps:
cd cafextractor/utils/OptExtract/
python setup.py build
mv build/lib.linux-x86_64-2.7/Marsh.so .
cd cafextractor/utils/CCF
./Proceso_f2py
Once installed, CAFExtractor is relatively simple to use. A simple example is:
- Place your data from night YYMMDD on a folder called:
mkdir /path_to_run_folder/00_RAW/YYMMDD- Go to CAFExtrcator source folder:
cd /path_to_cafextractor/cafextractor/src/- Run CAFExtractor:
python cafex.py YYMMDD --root /path_to_run_folder/The CAFExtractor package contains an "example" folder where you can test if the instalation was successful. To run the example just go to the cafextractor/src/ folder and do:
python cafex.py 190709 --root /path_to_cafextractor/examplesThis will create the folder /path_to_cafextractor/examples/11_REDUCED/190709 with the reduction of the science frames in the folder located in /path_to_cafextractor/examples/00_REDUCED/190709
RV re-computation
In some cases, you would like to recompute your RVs because of different reasons. The script RV_measure_onredspec.py on the src folder allows you to do this with different options like including the Simbad-searchable name of your object for a better computation of the BERV in case the OBJECT name put in the header is not resolved by Simbad. Alternatively, you can provide the coordinates of your target. Some examples follows:
python RV_measure_onredspec.py /path_to_cafextractor/examples/11_REDUCED/190709/reduced/HD109358__190709_0052_red.fits --COORD 12:33:44.54 +41:21:26.92 --RVguess 6.2 --RVampl 100. --UPDATERVFrom fits to ascii
The script fit2ascii_cafex.py also allows to convert the spactra into ascii files:
python fits2ascii_cafex.py /path_to_cafextractor/examples/11_REDUCED/190709/reduced/Reading cafeX spectra
The script read_cafex.py allows you to easily read the resulting extracted spectra and header from Python. Just do:
import read_cafex as rcafex
spec = rcafex.spec('/path_to_cafextractor/examples/11_REDUCED/190709/reduced/HD109358__190709_0052_red.fits')and you will get an object called "spec" that will contain the spec.wave, spec.flux, spec.eflux arrays that you can plot. For instance, if you would like to plot the 23rd order, just do:
plt.plot(spec.wave[23,:], spec.flux[23,:])
plt.show()Another option is to load a bunge of keyword parameters from the header of a set of files from the same object. A good example of why doing this is, for instance, if you have 20 spectra of a particular target and you want to see the RVs obtained on all spectra at once. Then just do:
import read_cafex as rcafex
obj = rcafex.obj('HD109358', MY_PATH='/path_to_cafextractor/examples/11_REDUCED/190709/reduced/')The object "obj" will then contain arrays like obj.rv or obj.hjd containing the RV and Julian Date (among others) values for all reduced files in the MY_PATH folder. Then you can just do:
plt.errorbar(obj.hjd,obj.rvcorr,yerr=obj.erv,fmt='o')- v0.1 01/2019 First release
- v0.2 01/2019 ThAr RV measurements changed
- v0.3 02/2019 New mask for science RV + fancy CCF plots
- v0.4 05/2019 New order geometry, new RV mask, RVCORR for SNR, new ref frames
- v0.5 05/2019 New RV corrections, new --root option, including CCF_FWHM in header
- v0.6 08/2019 Modified to reduced data from 2018
- v0.7 - -
- v0.8 11/2019 First version publicly released
- Order tracing: At high S/N there is a tilt of the extracted orders. To be investigated
- RV dependency with S/N: there is a strong dependency of RV with the S/N of the spectrum. Under investigation.
- Order normalization: Orders #X and #Y have a very bad normalization due to the strong atmospheric absorption at those wavelengths.
- Order merging: to be optimized
- Create IRAF-readable reduced spectra files (also MIDAS if possible)
If you make use of the products of the CAFExtractor pipeline please make sure to include the following reperences:
- Lillo-Box et al., 2020, MNRAS, 491, 4496 (https://ui.adsabs.harvard.edu/abs/2019arXiv190604060L/abstract)
- Brahm et al., 2017, PASP, 129, 973 (https://ui.adsabs.harvard.edu/abs/2017PASP..129c4002B/abstract)
A suggested sentence to include both references is as follows: "The data were reduced using the CAFExtractor pipeline \citep{lillo-box2019}, partly based on the CERES algorithms \citep{brahm2016}"
These tools have been developed by Dr. Jorge Lillo-Box. For any questions related to this pipeline please contact me at jlillo at cab.inta-csic.es .
I thank here the Calar Alto personnel for their great and outstanding support when preparing this pipeline. Also thanks to Arto Jarvinen (AIP) for helping with the installation details and feedback on the installation.