Reproducible Computational Experiments Using MADAGASCAR Software Package

1
Reproducible Computational Experiments Using
MADAGASCAR Software Package

• Sergey Fomel
• Bureau of Economic Geology
• University of Texas at Austin

Applied Inverse Problems Vancouver BC June 29,
2007
http//rsf.sf.net/
2
Principles of Scientific Software

• Encapsulation
• File Formats
• Testing
• Reproducibility
• Maintenance

3
Principles of Scientific Software

• Encapsulation
• File Formats
• Testing
• Reproducibility
• Maintenance

4
Encapsulation

• Information hiding (Parnas, 1972)
• Separation of concerns (Dijkstra, 1974)
• Separate physics from mathematics
• A is physics
• Going from b to is mathematics

5
Example Velocity Transform
6
Physics of Velocity Transform
7
(No Transcript)
8
(No Transcript)
9
Encapsulation in Programming

• Separation of concerns
• Classes or templates (C)
• Function pointers (C)
• Function interfaces (Fortran-90)

/ initialize velocity transform (A) /
veltran_init (true, x0, dx, nx, s0, ds, nv, o1,
d1, nt, s02, anti, psun1, psun2) /
least-squares minimization of A x b2,
xvscan, bcmp / sf_solver (veltran_lop,
sf_cgstep, ntv, ntx, vscan, cmp, niter,
"err", error, "nmem", 0, "nfreq", miter, "mwt",
mask, "end")
10
Encapsulation in UNIX

• Write programs that do one thing and do it well.
• Write programs to work together.
• Write programs to handle text streams, because
  that is a universal interface.

11
Encapsulation in UNIX Shell
bash sfveltran lt cmp.rsf gt vtran.rsf adjy v01
dv0.025 nv60 bash sfdottest sfveltran
modvtran.rsf datcmp.rsf v01 dv0.025
nv60 sfdottest Lmd21665.9 sfdottest
L'dm21665.9 bash sfdottest sfveltran
modvtran.rsf datcmp.rsf v01 dv0.025
nv60 sfdottest Lmd21906.2 sfdottest
L'dm21906.2 bash sfconjgrad sfveltran lt
cmp.rsf gt vtran.rsf niter3 v01 dv0.025 nv60
sfconjgrad iter 1 of 3 sfconjgrad
grad6.36797e09 sfconjgrad iter 2 of
3 sfconjgrad grad1.39068e09 sfconjgrad iter 3
of 3 sfconjgrad grad7.50257e08
12
Principles of Scientific Software

• Encapsulation
• File Formats
• Testing
• Reproducibility
• Maintenance

13
The Art of UNIX Programming

• (Raymond, 2004)
• To design a perfect anti-Unix, make all file
  formats binary and opaque, and require
  heavyweight tools to read and edit them.
• If you feel an urge to design a complex binary
  file format, or a complex binary application
  protocol, it is generally wise to lie down until
  the feeling passes.

14
RSF (Regularly Sampled Format)

• SEPlib (Stanford Exploration Project)
• Data separated from text headers
• Conceptually N-dimensional hypercubes
• Multiple files for complex geometries
• Not application specific

15
Principles of Scientific Software

• Encapsulation
• File Formats
• Testing
• Reproducibility
• Maintenance

16
Testing

• Test-driven development (Beck, 2003)
• YAGNI principle
• Always implement things when you actually need
  them, never when you just foresee that you need
  them.
• In scientific software development, tests are
  computational experiments

17
Testing with SCons

• Software Construction
• Replacement for make
• reliable and extensible dependency analysis
• configuration files are Python scripts
• cross-platform
• open-source

18
SConstruct File
Mobil AVO CMP gather 807 at well4
location Fetch('cmp807_raw.HH','rad')
Preprocessing Flow('cmp','cmp807_raw.HH',
'dd formnative tpow tpow2 mutter halfn
v01.3 tp0.2') Plot('cmp','grey title"Input CMP
Gather" ) Velocity Transform Flow('veltran','c
mp','veltran s020.25 v01.250 dv0.025 nv60
adjy') Plot('veltran','grey title"Velocity
Scan" ') Display Side by Side Result('veltran',
'cmp veltran','SideBySideAniso')
19
Experimenting with SCons
bash scons retrieve("cmp807_raw.HH", ) lt
cmp807_raw.HH sfdd formnative sftpow tpow2
sfmutter halfn v01.3 tp0.2 gt cmp.rsf lt
cmp.rsf sfgrey title"Input CMP Gather" gt
cmp.vpl lt cmp.rsf sfveltran s020.25 v01.250
dv0.025 nv60 adjy gt veltran.rsf lt veltran.rsf
sfgrey title"Velocity Scan" gt veltran.vpl vppen
yscale2 vpstylen gridnum2,1 cmp.vpl
veltran.vpl gt Fig/veltran.vpl bash sed
s/Velocity/Slowness/ lt SConstruct gt
SConstruct2 bash mv SConstruct2 SConstruct bash
scons lt veltran.rsf sfgrey titleSlowness Scan"
gt veltran.vpl vppen yscale2 vpstylen
gridnum2,1 cmp.vpl veltran.vpl gt Fig/veltran.vpl
20
Principles of Scientific Software

• Encapsulation
• File Formats
• Testing
• Reproducibility
• Maintenance

21
Reproducible Research at Stanford

• (Knuth, 1992)
• A computer program should be written with human
  readability as a primary goal.
• (Claerbout and Karrenbach, 1992)
• The purpose of reproducible research is to
  facilitate someone going a step further by
  changing something.
• (Buckheit and Donoho, 1995)
• An article about computational science in a
  scientific publication is not the scholarship
  itself, it is merely advertising of the
  scholarship.

22
Reproducible Experiments

• Within the world of science, computation is now
  rightly seen as a third vertex of a triangle
  complementing experiment and theory. However, as
  it is now often practiced, one can make a good
  case that computing is the last refuge of the
  scientific scoundrel Where else in science
  can one get away with publishing observations
  that are claimed to prove a theory or illustrate
  the success of a technique without having to give
  a careful description of the methods used, in
  sufficient detail that others can attempt to
  repeat the experiment? (LeVeque, 2006)

23
(No Transcript)
24
(No Transcript)
25
Principles of Scientific Software

• Encapsulation
• File Formats
• Testing
• Reproducibility
• Maintenance

26
Maintenance

• Computational experiments that are not
  continuously maintained loose reproducibility.
• Regression testing (Brooks, 1975)
• Contribute computational software and experiments
  to a community-maintained repository to enable
  research productivity.

27
Open Science
28
Conclusions

• Principles of Scientific Software
• Encapsulation
• File Formats
• Testing
• Reproducibility
• Maintenance
• Madagascar software package
• Open source, open community, open science