RV_PS2017

Jupyter Notebooks for the radial velocities tutorial at the Precision Spectroscopy Workshop 2017, held at the Instituto de Astronomia, Geofísica e Ciências Atmosféricas (Universidade de São Paulo) in August 2017. If you have suggestions or corrections, just open a GitHub issue.

Summary of this README:

1. Requirements
2. Setup
3. Technical background
4. Known issues

1. Requirements

In order to follow this tutorial, it is necessary to have basic knowledge of command line in Unix-based systems. Having experience with Jupyter notebooks (a.k.a. IPython notebooks) and knowing how to use git and conda is recommended. The Python packages pandas, radvel and radial (and their dependencies) need to be installed (see item 2). Basic knowledge of radial velocities and orbital parameters is also recommended (see item 3).

Also, make sure that pdflatex is installed. You can get pdflatex by installing the TexLive package or other LaTeX distributions. By default it is expected to be in your system’s path, but you may specify a path to pdflatex using the --latex-compiler option at the radvel report step.

2. Setup

We highly recommend using the Anaconda ecosystem, since it allows us to use the conda Python package and environment manager. Additionally, we recommend downloading the Python 3 version of Anaconda -- but do not worry about Python 2 codes, because we can easily setup a Python 2 environment using conda.

For the radial velocities tutorials, we will create a Python 2.7 environment called rv, and install the packages numpy, scipy, cython, astropy, pandas and matplotlib right from the start:

conda create -n rv python=2.7 numpy scipy cython astropy matplotlib pandas

Whenever you want to work in the rv environment, just issue the following command in the terminal:

source activate rv

And when you want to leave the environment, just use the following command:

source deactivate

Now, let's activate the rv environment and install some additional packages using the command pip:

source activate rv
pip install emcee corner lmfit pp jupyter

Note: jupyter can also be install using the command conda install jupyter.

Now we will proceed to install the radial radial velocities fitting codes. radvel is one of the RV packages we will use in this tutorial. It is authored by B. J. Fulton and E. Petigura. Unfortunately it works only in Python 2, which is why we setup the rv environment with python=2.7. The other package we will use in this tutorial is radial, which is authored by L. dos Santos (it works in both Python 2 and 3, but it's optimized for Python 3).

In order to install radial, go to the terminal and navigate to the folder you want to save the source code and issue the commands:

git clone https://github.com/RogueAstro/radial.git
cd radial
python setup.py develop

To install radvel, I recommend compiling the source code from the authors repository. In the terminal, navigate to the folder you want to save the source code and:

git clone https://github.com/California-Planet-Search/radvel.git
cd radvel
python setup.py develop

You can also clone the version of radvel that I forked from the original repository. I plan on modifying radvel in the near future to work with Python 3 and possibly adding some features here and there. Another option is to use pip install radvel, but I do not recommend it.

3. Technical background

The basic notions of radial velocities for binary stars and exoplanets can be found in chapter 2 of the book Exoplanets by Sara Seager. This chapter is authored by C. D. Murray and A. C. M. Correia, and it is freely available on arXiv.

Further reading for the enthusiasts:

• Minimum masses from minimal RV variation: Torres (1999), Feng et al. (2015), Jenkins et al. (2015)
• Alpha Cen Bb does not exist: Rajpaul et al. (2015)
• The radial velocity fitting challenge: Dumusque (2016)
• Systemic Console: Meschiari et al. (2009)
• EXOFAST: Eastman et al. (2013)
• Markov Chain Monte Carlo techniques: Ford (2005)
• Basic emcee example
• Fitting a plane to data
• A practical guide to the Lomb-Scargle Periodogram
• Frequentism vs. Bayesianism (Spoiler: Frequentism sucks)
• Toolkit for planet detection and characterization

4. Known issues

4.1. Compilation of C codes fails

You may run into problems when installing Python packages that compile C code with Anaconda. If the installation of the packages jupyter or radvel fails, use a custom installation of gcc. First, activate the rv environment if is not activated yet, and install the custom gcc:

source activate rv
conda install -c asmeurer gcc=4.8.5

This will install the custom gcc in the rv environment, but will leave the original installation of gcc untouched outside of this specific environment.

4.2. Jupyter notebooks do not initialize the right kernel

Have you tried turning it off and on again? No, seriously, if you installed jupyter inside the rv environment, you need to deactivate and then reactivate the rv environment again in order to initialize the correct kernel.

4.3. I get a NameError when running MCMC

When executing the mcmc part of radvel, you may run into the following error:

NameError: global name 'Pool' is not defined

To be honest, I don't exactly know what causes this issue at the moment. I got it in my Mac, but not in my Linux workstation, even though I followed the same setup in both machines. The only solutions I found is to either compile radvel from the source or install the Python 2 version of Anaconda and use that as the package manager.