Skip to content

b1quint/samfp-tools

Repository files navigation

SAM-FP

Build Status Coverage Status

This page contains some scripts that help me (and hopefully will help you too) to reduce Fabry-Perot data obtained at SOAR Telescope withSOAR Adaptive Module, or SAM.

This observing mode is considered a Restricted User Instrument and you will have support to reduce, edit, visualize and manipulate your data.

Scripts for data-reduction are saved inside the scripts folder. Each script is described individually in our Wiki page.

How to Install

We wrote samfp in Python 2 but, in principle, it should be compatible with Python 3. We simply did not have the chance yet to actually test the Py2 to Py3 conversion.

Since there are several ways which one can install a Python library, we decided to develop samfp using the Virtual Environment Astroconda under Anaconda. Once you have both installed and running properly, activate the Astroconda Virtual Environment by typing

 source activate astroconda

in a terminal.

After that, use pip to install samfp and its requirements:

cd $path_to_samfp
pip install -r requirements.txt
pip install . 

There is a know problem where both Numpy and SciPy fails to run under Anaconda because of a library called mkl. To fix that, type the following command

conda install nomkl

After that, you should be able to run the SAM-FP Tools scripts from anywhere in your system, as long as you are the Virtual Environment where they were installed (e.g. Astroconda).

Running the Scripts

xjoin

The images acquired with SAM-FP come from SAMI, the SAM Imager. This instrument provides FITS files with four extensions. We first combine all the extensions using the script xjoin. This will create a copy of the original file with a prefix xj_. You may add other options when combining the extensions like removing cosmic rays using LACosmic adding the -r or cleaning known bad columns when observing with binning mode 4 x 4 pixels with the -c flat. These flags are also added to the prefix.

BIAS images can also be subtracted from the image that is being processed using a -b flat and giving the name of the master bias file to the program.

FLAT corrections can also be applied by provinding the -f flag and the master flat filename.

combine_zero

TBD

combine_flat

TBD

phmxtractor

TBD

phmfit

TBD

phmapply

TBD

fpwcal

TBD

fpxmap

TBD