Introduction

PHY3033 – Advanced Astrophysics

Computer Lab

Philipp Edelmann
philipp.edelmann@ncl.ac.uk

Goals of this class

learn to run the stellar evolution code MESA
practical insight into stellar evolution
practical understanding of numerical simulations
learn how to analyse simulation output
learn to work in a Unix environment

The MESA stellar evolution code

solves equations of stellar structure
physics modules for nuclear reactions, opacities, …
can also simulate planets
free software (GPL)
developed by international community
on-the-fly visualisation
details in the next session

Equations of stellar structure

\begin{align} \PD{r}{m} &= \frac{1}{4 \pi r^2 \rho},\\ \PD{p}{m} &= - \frac{G m}{4 \pi r^2},\\ \PD{l}{m} &= \epsilon_\text{n} - \epsilon_\nu - c_p \PD{T}{t} + \frac{\delta}{\rho} \PD{p}{t},\\ \PD{T}{m} &= - \frac{G m T}{4 \pi r^4 p} \nabla,\\ \PD{X_i}{t} &= \frac{m_i}{\rho} \left(\sum_j r_{ji} - \sum_k r_{ik} \right). \end{align}

List of fastest supercomputers

top500.org

Rank	Site	System	Cores	Rmax (TFlop/s)	Rpeak (TFlop/s)	Power (kW)
1	National Supercomputing Center in Wuxi China	Sunway TaihuLight NRCPC	10,649,600	93,014.6	125,435.9	15,371
2	National Super Computer Center in Guangzhou China	Tianhe-2 (MilkyWay-2) NUDT	3,120,000	33,862.7	54,902.4	17,808
3	Swiss National Supercomputing Centre (CSCS) Switzerland	Piz Daint Cray Inc.	361,760	19,590.0	25,326.3	2,272
4	DOE/SC/Oak Ridge National Laboratory United States	Titan Cray Inc.	560,640	17,590.0	27,112.5	8,209
…
11	United Kingdom Meteorological Office United Kingdom	Cray Inc.	241,920	7,038.9	8,128.5	3,629

Top 500

Getting access to MESA

remotely log in to our Linux server
- no need to set anything up
- data analysis and plots can be done there or on another computer
- access to your Windows H: drive
install it on your own computer
- quite easy if you are running Linux or macOS
- on Windows it can be installed in a virtual machine
  (will be provided if there is interest)

Resources

exercise sheets, template work directories and slides on Blackboard
Gitlab server with updated template work directories:
https://mas-gitlab.ncl.ac.uk/PHY3033/exercises
introduction to Linux (by D. Robertson)
slides on Python for analysis/plotting

A very short introduction to Unix

family of operating systems sharing common code/design principles
developed in the 1970s at Bell Labs
portable: from phones to super computers
runs majority of internet servers
modular design (do one thing well)
many free variants: GNU/Linux, BSD, …

The command line

typically accessible from a terminal emulator
different shells with slightly different interface
(bash is most common)
used to string together several commands (pipes)
shells can be used to write scripts

File system tree

everything is in one tree (root at /)
all file systems are “mounted” at a path
directory separator: /
some useful directories:
- /home/username: home directory of the user
  can be abbreviated as ~
- /tmp: local temporary file storage for everyone, deleted regularly
- /etc: global configuration files
- /usr/bin: executables for installed software
- /media: mounted external file systems (e.g., USB drives)
- others mostly relevant for administration

Directories

see current working directory: pwd or in prompt
change directory: cd /tmp
list contents of directory: ls
extended listing: ls -l
current directory: . (e.g., cp /a .)
parent directory: .. (e.g., cd ..)

Environment variables

names associated with a character string
conventionally all upper case (not the content)
automatically passed along when running commands
can be used when constructing commands:
```
echo $USER
```
special variables: PATH, HOME, DISPLAY
set variables with
```
export MESA_DIR=$HOME/path/to/mesa
```

Running executables

executable name, optionally followed by one or more arguments (e.g., ls -l /tmp)
name is searched in PATH variable
entries separated by :
current directory (.) not part of PATH
run executables with ./myexec
full path can always be used
```
/bin/ls
```

Editing files

configuration often stored in plain text files
MESA is controlled through text files
MESA output is written to plain text files
simplest data format: one byte, one character (ASCII)
graphical editors: pluma, kate, mousepad, gedit
command-line editors:
vi, emacs, nano (easy for beginners)

Obtain exercise templates via git

git is a version-control system (VCS)
often used for (collaborative) software development
we distribute templates for exercises with it
alternative: zip file on Blackboard

get initial copy with

git clone https://mas-gitlab.ncl.ac.uk/PHY3033/exercises.git

update later with
```
git pull
```

Run a MESA session for testing

navigate to the template directory
exercises/00-test/work using cd
run the command ./mk to compile the example
run the code with ./rn
watch a $1\,M_\odot$ star evolve
you can interrupt it any time with Ctrl+c in the terminal