Matlab for a computational PhD

I M SE T
Matlab for your Computational PhD
pmxal9@nottingham.ac.uk
12 Oct 2016

I M SE T
Introduction
Matlab
Software Engineering
1

I M SE T
This talk
What this talk is about:
• Introducing the Matlab environment,
• Giving tips for a better use of Matlab,
• Giving pointers you can dig at home,
• Recommending Software Engineering practices.
What this talk is not about:
• Numerical schemes,
• Modelling,
• Big Data,
• Happiness and how is happens.
2

I M SE T
To Keep in Mind During a PhD (self-help++)
https://graduable.com/2012/10/04/
21-things-every-phd-student-should-know/
Take care of yourself, vision of Academia likely far from reality, learn
to read, it’s okay to change ... anything (project, supervisors),
depression, get to know the other postgrads, . . .
https:
//www.theguardian.com/higher-education-network/blog/2014/
may/02/five-things-successful-phd-students-refuse-to-do
Feel like a failure, feel out of control...
http://www.nextscientist.com/
graduate-school-advice-series-starting-phd/
Aim to publish peer-reviewed article, hard to stay motivated,
networking is important. . .
Be curious: Look for/after yourself :)
3

I M SE T
What is a PhD?
• Getting skills to perform research tasks,
• Become the expert on one topic,
• Start your career (make a network and publish),
• Move your research field forward (in practice, that’s often
long after the PhD, but act as if).
So why do good Software Engineering practices matter?
• Scientists spend > 30% of their time on coding [1],
• More than 90% are self-taught [1],
• Substantial quality problems,
• Increase of scientific errors due to incorrect Software [2],
• Opens up job/career opportunities, particularly in industry.
[1] P. Prabhu et al, Proc. 24th ACM/IEEE Conference on HPC, Networking, Storage and
Analysis (2011)
[2] Z. Merali, Error: Why scientific programming does not compute. Nature 467: 775777
(2010).
4

I M SE T
What is a PhD Thesis?
Umberto Eco takes us back to the original purpose of theses and
dissertations as deﬁning events that conclude a program of
study. They are not a test or an exam, nor should they be. [...]
Rather, they prove that students can make something out of
their education. This is particularly important today, when
we are more accustomed to thinking in compliance with the
software of our laptop or doing research according to the logic of
a tablet than to thinking and researching in a personal and
independent way.
The humanities are intrinsically creative and innovative. They
are about originality and invention, not discovery. This is
precisely Eco’s testimony; even more than a technical manual,
this book is an invitation to ingenuity, a tribute to
imagination.
Preface of ‘How to Write a Thesis’
5

I M SE T
Supervision
Or its absence. . .
http://academiaiskillingmyfriends.tumblr.com/
Most supervisors consider your programming/SE skills are your
problem. This is bound to change, but takes time... For now,
It’s your responsibility.
Partner up and go.
Another pitfall is to spend too much time on it. Research
remains your goal.
6

I M SE T
Best Practices for Scientific Computing
Scientists [...] lack exposure to basic software
development practices such as writing maintainable code,
using version control and issue trackers, code reviews, unit
testing, and task automation.
A large body of research has shown that code reviews are the
most cost-effective way of finding bugs in code. They are
also a good way to spread knowledge and good practices around
a team. In projects with shifting membership, such as most
academic labs, code reviews help ensure that critical knowledge
isn’t lost when a student or postdoc leaves the lab.
http://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.1001745
7

I M SE T
General Tips on your Research (1/3)
• Share your ideas,
• Let go of what does not work,
• Ask for advice after having given some thoughts
http://www.hbs.edu/faculty/Publication%20Files/Advice%20Seeking_
59ad2c42-54d6-4b32-8517-a99eeae0a45c.pdf
9

I M SE T
General Tips on your Computer (2/3)
Use Version Control
Bitbucket, Git. . . (bitbucket has private folders)
Bibliography Manager
Choose one, starting today: Zotero, Mendeley. . .
Notations
No space in folders & ﬁle names
Folder Organisation
Think carefully of a system & stick to it (until next version)
http://www.howtogeek.com/howto/15677/
zen-and-the-art-of-file-and-folder-organization/
10

I M SE T
Back-up
Your computer will crash and you’ll lose your data
LaTeX is a programming language
Modularity with input, make useful commands with
newcommand, make variables with def or edef for your
computer paths. . .
Command-line Editor
Learn to use the servers (Vim, Emacs. . . )
Unix
Learn Unix basics:
cat, grep, sed, find, ssh, scp, man, head, cd,
ls, rm, wc, xargs.
http://swcarpentry.github.io/shell-novice/)
11

I M SE T
Back-up
Buy a hard-drive, online (Dropbox, Bitbucket) or whatever.
Like, seriously...
12

I M SE T
Choosing Your Language
Mario Mulansky, ex-NETT, currently working for Apple
13

I M SE T
Matlab
Don’t use Matlab... but if you have to, know your tool!
Matlab...
is a nice working environment,
has a lot of useful toolboxes (and we have many: run ‘ver’)
is great for fast prototyping,
has OOP capabilities since 2008,
has some computational strengths,
is used in many Science ﬁelds and in industry,
is expensive (alternative: Octave),
has memory leaks,
is weak for text processing,
is easily slow,
...
14

I M SE T
How to use Matlab
• Command-line in Matlab environment
• Running scripts in Matlab command window
• Command-line in a shell (fastest)
$ matlab -nodisplay -nosplash < path/to/myscript.m
• Write GUIs 15

I M SE T
Adapt your Matlab Environment
Home > Layout
• Add column,
• Move layout here and there,
• Add/Remove Workspace, Command History,
• ...
16

I M SE T
Functions VS Scripts
Functions should do one thing only
and do it well. Their name should give this information, a help
describe their use and incorporate a minimal working example.
Example
cellText = importdata(’myfile.txt’);
firstValue = str2double(cellText{1});
17

I M SE T
Functions VS Scripts
Scripts should be your workﬂow
and clearly show what you do along the way.
Example
a = randn(100);
eigA = eig(a);
figure; plot(eigA, ’.’);
saveas(gcf, ’path/to/eigA.pdf’);
18

I M SE T
Workspaces
All the variables that are accessible within a function are
• either given as inputs to the function, or
• loaded or computed within the function.
Example
Within the function song2txt, just after the first line of code is
executed, only the variables wavfile, txtfile, wav and fs exist!
19

I M SE T
Workspaces
The Base workspace stores variables that you create at the
command line. This includes any variables that scripts create,
assuming that you run the script from the command line or
from the Editor.
Functions do not use the base workspace. Every function has its
own Function workspace. Each function workspace is
separate from the base workspace and all other workspaces to
protect the integrity of the data.
Example
What is the value of ‘a’ after we run
a = 0; myworkscript; myworkfunction();?
20

I M SE T
Sharing Data across Workspaces
In most cases, variables created within a function are local
variables known only within that function. But, since sharing is
caring. . .
• Passing arguments (best practice),
• Nested functions (not very popular),
• Evaluating in another workspace (bad practice),
• Saving data in a ﬁle and loading it later (slow),
• Persistent variables (very useful),
• Global variables (bad practice).
https://uk.mathworks.com/help/matlab/matlab_prog/
share-data-between-workspaces.html
21

I M SE T
22

I M SE T
23

I M SE T
24

I M SE T
25

I M SE T
A common way of passing many arguments with only a few
inputs is to use structures.
Example: myworkfunction2
26

I M SE T
Let’s compare with two other possibilities
27

I M SE T
Stack
Stack
When calling functions within functions, Matlab keeps track of
the order in which functions call other functions in the stack.
In debug mode, cDb = dbstack(); outputs a structure array.
You can also navigate it using the ‘Function Call Stack’.
Example
28

I M SE T
Debugging
Debugging takes time: help yourself by reducing it.
29

I M SE T
Debugging
Enter Debugging mode by setting Breakpoints within your code,
so that you can access and check all values at this point.
31

I M SE T
Debugging
In debug mode, you can see the function variables within the
‘Worspace’ tab, and read variables values by hovering over them.
32

I M SE T
Debugging
Debugging tip
Use dbstop if error during development to enter debug mode
instead of crashing. Don’t use in general because most of the
time you know what the error is by reading the error message.
Disable with dbclear if error.
33

I M SE T
Numeric Types
Numeric Types
Double, single, int8, int16, int32, int64, uint8, . . .
Big Data? Bottlenecks? Depending on your needs, some
numeric types could be better suited than others.
34

I M SE T
Main Data Classes
Array a = [1, -2.5]; b = randn(3); c = ’hi’;
Cell a = {1, [1 2], {1 2 3}, ’hi’};
Structure a.f1 = 1; a.f2 = ’2’; a.f3 = {4};
Struct. array a(2).f1 = 0;
Table a = array2table([1 2],’Var’,{’x’ ’y’})
Categorical a = categorical([3 3 4 4])
function_handle a = @(x)2*x
(OOP) a = albanissocurly;
(Graphical) a = figure;
Functions class, size and isa useful for automation.
35

I M SE T
Structure Array VS Tables
Structure arrays are more common among Matlab users than
Tables, but the latter present a certain number of advantages:
Stuct. Array Table
Extracting column of
numbers/strings takes
two semantics:
a1 = [a.dou];
a2 = {a.str};
one semantic:
a1 = a.dou;
a2 = a.str;
Switch to categories to
reduce redundancy is
not easy
easy:
a.str =
categorical(a.str)
Having a summary of
the information is
not easy
easy:
summary(a)
36

I M SE T
Example: » summary(T)
Description: Simulated Patient Data
Variables:
Gender: 100x1 cell string
Description: Male or Female
Age: 100x1 double
Units: Yrs
Values:
min 25
median 39
max 50
Smoker: 100x1 logical
Description: true or false
Values:
true 34
false 66 37

I M SE T
Example: Reduce redundancy
If your table contains a lot of redundancy, using categories can
be very time- and memory-eﬃcient. Requires care. 38

I M SE T
Take-home message
If you need to create a lot of complex data (Data Wrangling),
think of the simplest table you can make to contain it. The
redundancy this entails can often be dealt with by using
categories.
Example
myData(5).meta(19).metaMeta(4).damnThisIsACell{4} = 1;
myTable.ThisIsACell(4) = 1;
https://uk.mathworks.com/help/matlab/ref/table.html
39

I M SE T
Data Exploration
Run in Matlab Command Window
mystruct.data = [linspace(0,20,100)’, rand(100,1)];
mystruct.description = ’Description’;
open mystruct
Click on ‘data’ ﬁeld
Click on PLOTS > pie
Repeat on a column of ‘data’
40

I M SE T
Loading & Saving Data
Three options for saving:
• Save Matlab variables in a .mat file to reuse them later
Simplest
Matlab only
• Save in a text file
Readable output (universal interface)
Can be used by all programming languages
Can use Bash utilities
Requires some planning ahead: appropriate format?
• Save in a binary file
Minimal size
Takes more programming time
Requires a lot of planning ahead: optimised format?
Non-readable output, hence much harder to debug
41

I M SE T
Loading data:
• Matlab variables: load
• Audio: audioread
• Images: imread
• Binary ﬁle: fscanf, fgetl, fgets, fread
• Text: textscan, csvread, dlmread, readtable
The import tool (Home > Variable > Import Data) and
importdata are quite general, but less robust.
42

I M SE T
Use Importdata to import myﬁle.txt into a structure
>> type(’myfile.txt’)
Day1 Day2 Day3 Day4
95.01 76.21 61.54 40.57
23.11 45.65 79.19 93.55
60.68 1.85 92.18 91.69
48.60 82.14 73.82 41.03
89.13 44.47 17.63 89.36
>> M = importdata(’myfile.txt’, ’ ’, 1);
>> disp(M);
M =
data: [5x4 double]
textdata: {’Day1’ ’Day2’ ’Day3’ ’Day4’}
colheaders: {’Day1’ ’Day2’ ’Day3’ ’Day4’}
43

I M SE T
Saving Plots
Automate saving
Plots are quite important;
Freezing how they were made is underrated,
Keep .ﬁg to avoid having to recreate it when changing the
format.
Example
set(gcf, ’PaperSize’, [12 4], ’PaperPosition’, [0 0 12 4]);
soundPlots = sprintf(’/Users/exp1058/%s’,func2str(func));
saveas(gcf, [soundPlots ’.pdf’]);
saveas(gcf, [soundPlots ’.fig’]);
44

I M SE T
Proﬁling
Proﬁling to increase performance comes last:
Make it work. Make it right. Make it fast.
Main Tools
• tic, toc
• profile on, profile viewer, profile off,
• Home > Code > Run and Time.
45

I M SE T
Proﬁling
or from tictocs.m
47

I M SE T
Regular expressions
Versatile way to search and replace text using patterns.
• Horrible to master but tremendously powerful,
• Understand them before the day you need them,
• Bash has more tools (cat, grep, awk, sed, tr. . . ),
• Matlab has a few functions: strfind, regexp, regexprep.
48

I M SE T
Conditions
if... elseif... else... end
Use if you have conditions that are apparently not related.
switch... case ... otherwise... end
Use for matching among some values (strings or numerical).
Example switch
swich class(f)
case ’cell’, disp(’This is a cell’);
case ’double’, disp(’This is an array’);
case {’you’, ’yourmom’}, disp(’is so fat’);
otherwise, error(’Not cool, man’);
end
49

I M SE T
Loops
for... end
Use when you have an index that you control.
Always allocate memory by deﬁning variables before the loop.
while... end
When for is not an option.
break
To terminate execution of a loop.
continue
To pass control to next iteration in a loop.
return
To exit a function when its job is done.
50

I M SE T
Regular expressions
Example in Matlab
labtext = importdata(labTXT);
labtext1 = strrep(labtext{1}, ’.’, ’’);
labtext2 = regexprep(labtext1, ’ds*’, ’’);
splitted = regexp(labtext2,’s+’, ’split’);
upsplt = upper(sprintf(’%sn’,splitted{:}));
upsped = regexprep(upsplt,’[0-9]|?|,|.|!|,|;|-|:’, ’’);
51

I M SE T
Regular expressions
Example in Matlab
labtext = importdata(labTXT);
% Get rid of dots, for consistency.
labtext1 = strrep(labtext{1}, ’.’, ’’);
% Get rid of initial numbers
labtext2 = regexprep(labtext1, ’ds*’, ’’);
% To write label in words.mlf; one word per line
splitted = regexp(labtext2,’s+’, ’split’);
% uppercase
upsplt = upper(sprintf(’%sn’,splitted{:}));
% Keep only letters
upsped = regexprep(upsplt,’[0-9]|?|,|.|!|,|;|-|:’, ’’);
52

I M SE T
Regular expressions
Example in Bash
#
pdftotext Thesis.pdf - | grep -o "[?" | wc -l
53

I M SE T
Regular expressions
Example in Bash
# Approx. count number of missing citations in Thesis.pdf
pdftotext Thesis.pdf - | grep -o "[?" | wc -l
54

I M SE T
Interfacing with Matlab
“Controlling complexity is the essence of computer
programming”
Brian Kernighan
→ If you can stick with one language only, do it!
Reasons to use other languages within Matlab:
• Reduce runtime
(‘Oh, this loop would be much faster in C...’)
• Use existing packages
(‘Oh, if only there was a way to use this .jar?’)
• Reduce development time
(‘Oh, I know how to do this in Bash!’)
55

I M SE T
Interfacing with Matlab: C with mex-files
• Very hard to debug (Matlab will crash often),
• Requires specific commands,
• Might boost your code,
• Takes significant development time.
56

I M SE T
Interfacing with Matlab: C with mex-ﬁles
57

I M SE T
Interfacing with Matlab: Java with jar-ﬁles
• Need to know exactly what
methods/classes you need,
• Hard to debug,
• Juggling between Java and
Matlab objects,
• Will develop your insight of
Matlab’s inner working.
58

I M SE T
Interfacing with Matlab: Bash with !
On a Linux machine, you can execute bash commands using ‘!’
Ideal for some quick-n-dirty tricks
59

I M SE T
Interfacing with Matlab: LATEXwith fprintf...
60

I M SE T
Programming Paradigms (Wikipedia)
61

I M SE T
Programming Paradigms in Matlab
Matlab is (weakly) multi-paradigm:
• Imperative Programming: Change program’s state with
commands describing how.
• Procedural programming: Procedures, also known as
routines, subroutines, or functions contain a series of
computational steps to be carried out.
• Object-Oriented Programming: Based on objects and
their methods and attributes that interact with each other.
62

I M SE T
Programming Paradigms in Matlab
Complex matter:
[...] Matlab pretends to be multi-paradigm. It is optimized for
imperative and procedural strongly typed operations. It
supports OO, and extends to C/Cpp and Java. Contrary to
popular opinion, Matlab is not strictly an interpreted language
and does not always use JIT compilation. It uses an opaque
optimization scheme [...] https://www.quora.com/
What-programming-paradigm-does-MATLAB-follow
63

I M SE T
Object-Oriented Programming in Matlab
Great if you work on big projects/with other people!
https://uk.mathworks.com/company/newsletters/articles/
introduction-to-object-oriented-programming-in-matlab.
html
64

I M SE T
Add-On Exploration
Add-Ons > Get Add-Ons
Search and install ‘Megan Simulator’
Simpler and faster than doing through your browser:
Installed in the right place,
Easy to manage,
Directly added to your path,
All gathered in one place,
You can write your own.
65

I M SE T
Boosting Matlab Performance
Parallelisation instead of for loops
Logical operator
Preallocation
Master COW (Copy On Write)
Parallel processing toolbox
Removing bottlenecks after proﬁling
. . .
https://uk.mathworks.com/help/matlab/matlab_prog/
techniques-for-improving-performance.html
66

I M SE T
Testing
If performance is a constraint/limitation, use a proper Testing
framework.
Matlab oﬀers a series of functions/classes/scripts to measure the
performance of your program
Special cases should become test cases.
http://uk.mathworks.com/help/matlab/performance-testing-framework.html
67

I M SE T
Literate Computing
Interactive document that combines MATLAB code with
embedded output, formatted text, equations, and images in a
single environment called the Live Editor.
https://uk.mathworks.com/help/matlab/matlab_prog/what-is-a-live-script.html
69

I M SE T
Unveiling Matlab Secrets
Yair Altman:
- Book ‘Accelerating MATLAB Performance’
- Awesome website http://undocumentedmatlab.com/
Check help/doc of all functions you use frequently.
Look at people’s code.
Google and ﬁddle
70

I M SE T
Software Engineering
‘The application of a systematic, disciplined, quantiﬁable
approach to the development, operation,
and maintenance of software.’
IEEE Standard Glossary of Software Engineering Terminology
Programming: Producing code that works correctly.
Software Engineering additionally:
• Readability, Maintainability, Documentation
• Unit Testing
• Design
• Code management, Task automation
71

I M SE T
Readability
Write programs for people, not computers.
• Case sensitive names (weight1 rather than param1),
• Comment (whenever code is not obvious),
• Give MWE (Minimal Working Example),
• Do not recode (use available software),
• Do not recode (small functions rather than copy-paste),
• Write short functions that do one thing.
72

I M SE T
Automation
• Use scripts (Matlab or bash) to execute code,
• Keep logs
Example: Bash → Matlab → log
matlab -nodisplay -nosplash < myscript.m > mylog.txt
73

I M SE T
Agile Development
• Work in small increments,
• Use version control where you save everything,
• Measure your progress,
• Use it to rethink how you can improve your own eﬃciency.
75

I M SE T
Agile Development
Make it work. Make it right. Make it fast.
76

I M SE T
Open: Source, Code, Data
Share your code, for ex. on GitHub; choose your licenses.
Increasingly important: we start to be evaluated by our
programming/SE skills.
Institutions (universities and governements) are being pushed to
share their code and data.
A code you can’t share is a code that is wrong.
Alban, 12th of October 2016, 2:18am
77

I M SE T
Summary of Best Practices
Readability Automation
Agile Programming DRY Principle
Correctness Optimisation
Documentation Collaboration
78

I M SE T
Resources speciﬁc to PhDs Nottingham
High-Performance Computing facility (HPC),
Maths servers (Shrek, Pegasus,...),
Nottingham PhD students,
Organise you own seminar :)
79

I M SE T
Software Engineering Resources
Greg Wilson
http://software-carpentry.org/lessons/
Best Practices for Scientiﬁc Computing, PLoS Biology 2014.
http://journals.plos.org/plosbiology/article?id=10.1371/
journal.pbio.1001745
Get some apps for your daily learning
https://medium.com/the-mission/
9-places-to-learn-how-to-code-in-15-minutes-or-less-a-day-7eb730e4fc82#.7gxej497p
80

I M SE T
Favourite Bash One-Liner from this PhD
To connect as pmxlol to the pegasus server,
reach the existing and detached screen session myscreen,
launch Matlab within it,
execute the script myscript.m,
save output in a log called mylog.txt:
ssh pmxlol@pegasus.maths.nottingham.ac.uk
"screen -S myscreen -X stuff
"matlab -nodisplay -nosplash < myscript.m > mylog.txt
$(echo -ne ’r’)""
Links to explain the diﬀerent bits:
https://code.tutsplus.com/tutorials/ssh-what-and-how--net-25138
https://www.gnu.org/software/screen/manual/screen.html
http://unix.stackexchange.com/questions/13953/sending-text-input-to-a-detached-screen
http://stackoverflow.com/questions/33187141/how-to-call-matlab-script-from-command-line
http://www.tldp.org/ldp/abs/html/io-redirection.html
81

I M SE T
Master Foo and the Recruiter
A technical recruiter, having discovered that that the ways of Unix hackers
were strange to him, sought an audience with Master Foo to learn more
about the Way. Master Foo met the recruiter in the HR offices of a large
firm.
The recruiter said, "I have observed that Unix hackers scowl or become
annoyed when I ask them how many years of experience they have in a new
programming language. Why is this so?"
Master Foo stood, and began to pace across the office floor. The recruiter
was puzzled, and asked "What are you doing?"
"I am learning to walk," replied Master Foo.
"I saw you walk through that door" the recruiter exclaimed, "and you are
not stumbling over your own feet. Obviously you already know how to
walk."
"Yes, but this floor is new to me." replied Master Foo.
Upon hearing this, the recruiter was enlightened.
82

I M SE T
References
Randall Munroe’s
https://xkcd.com/
Eric Steven Raymond’s
http://catb.org/esr/writings/unix-koans/index.html
Unix Philosophy
http://www.catb.org/esr/writings/taoup/html/index.html
Mario Mulansky’s ‘Programming for Scientists’
Best Practices for Scientiﬁc Computing, PLoS Biology 2014.
http://journals.plos.org/plosbiology/article?id=10.
1371/journal.pbio.1001745
83

I M SE T
Thanks for your attention!
Merci pour votre attention !
84

I M SE T
Master Foo and the Ten Thousand Lines
Master Foo once said to a visiting programmer: "There is more Unix-nature in one l ine of
shell script than there is in ten thousand lines of C."
The programmer, who was very proud of his mastery of C, said: "How can this be? C is the
language in which the very kernel of Unix is implemented!"
Master Foo replied: "That is so. Nevertheless, there is more Unix-nature in one line of shell
script than there is in ten thousand lines of C."
The programmer grew distressed. "But through the C language we experience the
enlightenment of the Patriarch Ritchie! We become as one with the operating system and the
machine, reaping matchless performance!"
Master Foo replied: "All that you say is true. But there is still more Unix-nature in one line of
shell script than there is in ten thousand lines of C."
The programmer scoﬀed at Master Foo and rose to depart. But Master Foo nodded to his
student Nubi, who wrote a line of shell script on a nearby whiteboard, and said: "Master
programmer, consider this pipeline. Implemented in pure C, would it not span ten thousand
lines?"
The programmer muttered through his beard, contemplating what Nubi had written. Finally
he agreed that it was so.
"And how many hours would you require to implement and debug that C program?" asked
Nubi.
"Many," admitted the visiting programmer. "But only a fool would spend the time to do that
when so many more worthy tasks await him."
"And who better understands the Unix-nature?" Master Foo asked. "Is it he who writes the ten
thousand lines, or he who, perceiving the emptiness of the task, gains merit by not coding?"
Upon hearing this, the programmer was enlightened. 86

Matlab for a computational PhD

More Related Content

Similar to Matlab for a computational PhD (20)

More from AlbanLevy (7)

Recently uploaded (20)

Matlab for a computational PhD