Beyond the Mouse A Short Course on Programming - 1 ... · Beyond the Mouse – A Short Course on Programming 1. Thinking programs Ronni Grapenthin Geophysical Institute, University

Beyond the Mouse – AShort Course on

Programming1. Thinking programs

Ronni Grapenthin

Geophysical Institute, University of AlaskaFairbanks

September 12, 2011 “The Uncomfortable Truths Well”,http://xkcd.com/568 (April 13, 2009)

1 / 21

Outline

1 Overview and Philosophies

2 Thinking programs

3 Building programs

4 Summary

2 / 21

Outline


2 Thinking programs

3 Building programs

4 Summary

3 / 21

The Program . . .

Rethink!NO!Need to

program?START

Listing 1: Class flow

4 / 21

The Program . . .

N = 1

Rethink!NO!

I guess

Need to program?

START


4 / 21

The Program . . .

N = 1

Rethink!NO!

I guess

Need to program?

START

Attend class N

N=1: "Introduction"

N=2: "Thinking Programs"

N=3: "Fundamental Principles I"

N=4: "Matlab I"

N=5: "Fundamental Principles II"

...

N=14: "GMT II"


4 / 21

The Program . . .

N = 1

Rethink!NO!

I guess

Need to program?

START

Attend class N

Interesting?Useful?

Accessible?

Yes.Play with examples

Send email with N and comments to Ronni

NO!

N=1: "Introduction"



N=4: "Matlab I"


...

N=14: "GMT II"


4 / 21

The Program . . .

N = 1

Rethink!NO!

I guess

Need to program?

START

Attend class N

Interesting?Useful?

Accessible?


Yes.Interesting?Useful?

Go through exercisesSend email with N and

comments to Ronni

NO!

NO!

N=1: "Introduction"



N=4: "Matlab I"


...

N=14: "GMT II"


4 / 21

The Program . . .

N = 1

Rethink!NO!

I guess

Need to program?

START

Attend class N

Interesting?Useful?

Accessible?




comments to Ronni

NO!

NO!

Have a l i fe!

N=1: "Introduction"



N=4: "Matlab I"


...

N=14: "GMT II"


4 / 21

The Program . . .

N = 1

Rethink!NO!

I guess

Need to program?

START

Attend class N

Interesting?Useful?

Accessible?




comments to Ronni

NO!

NO!

Have a l i fe!

N=1: "Introduction"



N=4: "Matlab I"


...

N=14: "GMT II"

Attend LAB N


4 / 21

The Program . . .

N = 1

Rethink!NO!

I guess

Need to program?

START

Attend class N

Interesting?Useful?

Accessible?




comments to Ronni

NO!

NO!

Have a l i fe!

N = N+1

N=1: "Introduction"



N=4: "Matlab I"


...

N=14: "GMT II"

Attend LAB N

Interesting?Useful?

NO! Send email with LAB N and comments to Ronni

Yes.


4 / 21

The Program . . .

N = 1

Rethink!NO!

I guess

Need to program?

START

Attend class N

Interesting?Useful?

Accessible?




comments to Ronni

NO!

NO!

Have a l i fe!

N = N+1

N < 14?

Yes.

N=1: "Introduction"



N=4: "Matlab I"


...

N=14: "GMT II"

Attend LAB N

Interesting?Useful?


Yes.


4 / 21

The Program . . .

N = 1

Rethink!NO!

I guess

Need to program?

START

Attend class N

Interesting?Useful?

Accessible?




comments to Ronni

NO!

NO!

Have a l i fe!

N = N+1

N < 14?

Yes.

NO! Send email with comments to Ronni

N=1: "Introduction"



N=4: "Matlab I"


...

N=14: "GMT II"

Attend LAB N

Interesting?Useful?


Yes.


4 / 21

The Program . . .

N = 1

Rethink!NO!

I guess

Need to program?

START

Attend class N

Interesting?Useful?

Accessible?




comments to Ronni

NO!

NO!

Have a l i fe!

N = N+1

N < 14?

Yes.

NO! Send email with comments to Ronni

N=1: "Introduction"



N=4: "Matlab I"


...

N=14: "GMT II"

Attend LAB N

Interesting?Useful?


Yes.

Exercise #1: What are possib le problems wi th

this program?

Think about t iming, count ing, and general f low.


4 / 21

The very basics (1)

From ‘The Conscience of a Hacker’, The Mentor (1986):[. . . ] I made a discovery today. I found a computer. Wait a second,this is cool. It does what I want it to. If it makes a mistake, it’sbecause I screwed it up. Not because it doesn’t like me . . .Or feels threatened by me . . .Or thinks I’m a smart ass . . .Or doesn’t like teaching and shouldn’t be here [. . . ]

5 / 21

The very basics (2)

Programming is beyond language.

Programming is about writing code that people can read.Code is poetry.RTFM and/or the internet

http://thinkgeek.com

“When I’m writing poetry,it feels like the center ofmy thinking is in aparticular place, andwhen I’m writing code thecenter of my thinking feelsin the same kind of place.”

Richard Gabriel,

Distinguished Engineer at Sun Microsystems

6 / 21

The very basics (2)

Programming is beyond language.Programming is about writing code that people can read.

Code is poetry.RTFM and/or the internet



Richard Gabriel,


6 / 21

The very basics (2)

Programming is beyond language.Programming is about writing code that people can read.Code is poetry.

RTFM and/or the internet



Richard Gabriel,


6 / 21

The very basics (2)

Programming is beyond language.Programming is about writing code that people can read.Code is poetry.RTFM and/or the internet



Richard Gabriel,


6 / 21

More Philosophy . . .

Jon Claerbout (a geophysicist), as quoted in “WaveLab andReproducible Research”:

An article about computational science in a scientificpublication is not the scholarship itself, it is merelyadvertising of the scholarship. The actual scholarship is thecomplete software development environment and thecomplete set of instructions which generated the figures.

Implications . . .publications should include data and code (example: Okada)figures should be reproducible by readerswrite code that others can use!

7 / 21

What does that mean?

Good1 function fp = screw2d ( x , xf , d , sdot )

% f u n c t i o n fp = screw2d ( x , xf , d , sdot )3 %

% Computes f a u l t−p a r a l l e l s l i p ra te f o r 2D screw d i s l o c a t i o n5 % wi th f a u l t loca ted at xf , w i th l ock ing depth d and s l i p ra te sdot .

% W i l l compute a t one or many l o c a t i o n s x .7 %

% x column vec to r9 % x f sca la r

% d sca la r11 % sdot sca la r

%13 i f ( d == 0 )

fp = sdot∗0.5∗sign ( x−x f∗ones ( size ( x ) ) ) ;15 else

fp = sdot∗atan2 ( ( x−x f∗ones ( size ( x ) ) ) , d ) / pi ;17 end

Badfunction fp = screw2d ( x , xf , d , sdot )

2 i f ( d==0) fp=sdot∗0.5∗sign ( x−x f∗ones ( size ( x ) ) ) ; else fp=sdot∗atan2 ( ( x−x f∗ones ( size ( x ) ) ) , d ) / pi ;end

8 / 21

Outline


2 Thinking programs

3 Building programs

4 Summary

9 / 21

Thinking programs – Breaking down complex tasks

Example 1:

Getting into grad school ... andout.

things to do:apply, figure out where to go, visa stuff, class work, research, thesis . . .

10 / 21


Find schools

Listing 1: Seminar flow

11 / 21


Find schools

search web

talk to advisors


11 / 21


Find schools

apply

search web

talk to advisors

GRE, TOEFL

paperwork


11 / 21


Find schools

apply

wait ...

search web

talk to advisors

GRE, TOEFL

paperwork


11 / 21


Find schools

apply

wait ...

make decision

search web

talk to advisors

GRE, TOEFL

paperwork


11 / 21


Find schools

apply

wait ...

make decision

move

search web

talk to advisors

GRE, TOEFL

paperwork

do visa stuff

pack and leave


11 / 21


Find schools

apply

wait ...

make decision

move

do classwork do research write papers

search web

talk to advisors

GRE, TOEFL

paperwork

do visa stuff

pack and leave


11 / 21


Find schools

apply

wait ...

make decision

move


write thesis

defend

search web

talk to advisors

GRE, TOEFL

paperwork

do visa stuff

pack and leave


11 / 21


Find schools

apply

wait ...

make decision

move


write thesis

defend

search web

talk to advisors

GRE, TOEFL

paperwork

do visa stuff

pack and leave

find a job


11 / 21


Example 2:

Grad student’s Average Day

possible activities:eat, sleep, work, do stuff, . . .

12 / 21


Example 2:

Grad student’s Average Daypossible activities:

eat, sleep, work, do stuff, . . .

12 / 21


get up

do morning stuff

go to school

work

lunch

work

dinner

work

go home

sleep

Listing 1: make_my_day

possible implementation1 % make_my_day .m

%−−−−−−−−−−−−−−3 % program t h a t shows how much fun

% l i v e as a grad student i s : )5

clc ;7

getUp ;9 eat ( ’ b reak fas t ’ ) ;

walk ( ’ school ’ ) ;11 work ;

eat ( ’ lunch ’ ) ;13 work ( ) ;

eat ( ’ d inner ’ ) ;15 work ( ) ;

walk ( ’home ’ ) ;17 haveLi fe ;

s leep ;

13 / 21


get up

do morning stuff

go to school

work

lunch

work

dinner

work

go home

sleep

Listing 1: make_my_day

possible implementation% make_my_day .m

2 %−−−−−−−−−−−−−−% program t h a t shows how much fun

4 % l i v e as a grad student i s : )

6 clc ;

8 getUp ;eat ( ’ b reak fas t ’ ) ;

10 walk ( ’ school ’ ) ;work ;

12 eat ( ’ lunch ’ ) ;work ( ) ;

14 eat ( ’ d inner ’ ) ;work ( ) ;

16 walk ( ’home ’ ) ;haveLi fe ;

18 sleep ;

13 / 21

Outline


2 Thinking programs

3 Building programs

4 Summary

14 / 21

Building programs – One black box at a time

Strategies to implement a program:

Top downSame as the examples above:

start with the big pictureidentify reasonablesubtaskstry to divide things to a levelof managable complexity(atoms)implement atomsimplement main routine(flow control)

Bottom upproblems accumulateimplement an atom at thetimeat some point you figure outthat things could gotogetherrevise main routineconstantlyadd necessary subroutines

15 / 21


Bottom lineTry building tools that solve a set of similar problems in a genericway. Use Parameters!Build and test each atom individually, test all scenarios (and more)with synthetic input.Treat atoms as black boxes that implement desired functionality.Don’t care about them once they’re working

16 / 21


Keys to good programsModularity: split problem in manageable tasks, implement andtest one at a time

Reusability: write functions, avoid redundance, avoid monolithiccode (theoretically one loop would be enough)

Pidd, 2002

Generalize: use variables instead of hard coded values, handparameters to functionsFunctionality, then efficiency

17 / 21


Keys to good programsModularity: split problem in manageable tasks, implement andtest one at a timeReusability: write functions, avoid redundance, avoid monolithiccode (theoretically one loop would be enough)

Full (composable) simulation model reuse

Function reuse

Copy ’n’ paste

Frequency:

Complexity:

low

lowhigh

high

Software com- ponent reuse

Pidd, 2002


17 / 21




Function reuse

Copy ’n’ paste

Frequency:

Complexity:

low

lowhigh

high


Pidd, 2002

Generalize: use variables instead of hard coded values, handparameters to functions

Functionality, then efficiency

17 / 21




Function reuse

Copy ’n’ paste

Frequency:

Complexity:

low

lowhigh

high


Pidd, 2002


17 / 21

Building programs

The Control Routine% make_my_day .m

2 %−−−−−−−−−−−−−−% program t h a t shows how much fun

4 % l i v e as a grad student i s : )

6 clc ;

8 getUp ;eat ( ’ b reak fas t ’ ) ;

10 walk ( ’ school ’ ) ;work ;

12 eat ( ’ lunch ’ ) ;work ( ) ;

14 eat ( ’ d inner ’ ) ;work ( ) ;

16 walk ( ’home ’ ) ;haveLi fe ;

18 sleep ;

Using Parameters% eat .m

2 %−−−−−−−−−−−−−−function eat ( what )

4 f p r i n t f (1 , ’%s : yummy . . . %s \ n ’ , . . .mfilename , what ) ;

6 pause ( 1 ) ;end

18 / 21

Outline


2 Thinking programs

3 Building programs

4 Summary

19 / 21

Summary – Take home messages

Thinking . . .Think modularThink in general casesThink non-redundantThink about reuseThink about reproducibility

Exercising . . .Read other peoples’ code . . . criticallyThe first version is for the trash bin (unintentionally)

Truth . . .Your working environment will change, concepts likely survive! Beflexible in the choice of languages and tools.

20 / 21

If all fails . . .

“The Ballmer Peak”

http://www.xkcd.com/323/

21 / 21

Overview and PhilosophiesThinking programsBuilding programsSummary

Documents

Beyond the Mouse A Short Course on Programming - 1 ... · Beyond the Mouse – A Short Course on Programming 1. Thinking programs Ronni Grapenthin Geophysical Institute, University