Code: Difference between revisions

From Octave
Jump to navigation Jump to search
(Statistics)
(Update description of dynare)
 
(9 intermediate revisions by 8 users not shown)
Line 2: Line 2:
*Francesco Potortì's fractal programs (Octave, GPL): http://fly.cnuce.cnr.it/software/#fractals
*Francesco Potortì's fractal programs (Octave, GPL): http://fly.cnuce.cnr.it/software/#fractals


==Ecnometrics==
==Econometrics==
*Michael Creel's Econometrics notes and octave files: http://pareto.uab.es/mcreel/Econometrics/ (try the Econometrics/MyOctaveFiles directory too)
*Michael Creel's Econometrics notes and octave files: http://pareto.uab.es/mcreel/Econometrics/ (try the Econometrics/MyOctaveFiles directory too)
* Ludwig Kanzler's Econometric software (much of it adapted to Octave):
* Ludwig Kanzler's Econometric software (much of it adapted to Octave):
**http://www2.gol.com/users/kanzler/index.htm#L.%20Kanzler:%20Software
**http://www2.gol.com/users/kanzler/index.htm#L.%20Kanzler:%20Software
**Licence: "All rights reserved", Unfree for distributing modified versions
**Licence: "All rights reserved", Unfree for distributing modified versions
*[http://www.dynare.org Dynare] is a software platform for handling a wide class of economic models, in particular dynamic stochastic general equilibrium (DSGE) and overlapping generations (OLG) models. Licensed under GNU GPL (v3 or later). Consists of a pre-processor (in C++) and of a collection of Octave/Matlab routines.


==Communications Toolbox==
==Communications Toolbox==
Line 24: Line 25:
*Simulate multipath fading channels: http://zeidler.ucsd.edu/~sct/code/20060112_multipath/
*Simulate multipath fading channels: http://zeidler.ucsd.edu/~sct/code/20060112_multipath/
*Signal processing package in [http://octave.sourceforge.net/signal/index.html OctaveForge]
*Signal processing package in [http://octave.sourceforge.net/signal/index.html OctaveForge]
*Wavelets: try Wavelab developed by David Donoho et al. at http://www-stat.stanford.edu/~wavelab/. For more information, see [[Wavelab on Octave]].
*Wavelets: LTFAT package in [http://octave.sourceforge.net/ltfat/ OctaveForge], see additional documentation in the [http://ltfat.sourceforge.net/doc/wavelets/index.php LTFAT web page]. Try also Wavelab developed by David Donoho et al. at http://www-stat.stanford.edu/~wavelab/. For more information, see [[Wavelab on Octave]].
* Peter Kovesi's functions for computer vision and image processing: http://www.csse.uwa.edu.au/~pk/Research/MatlabFns/
* Peter Kovesi's functions for computer vision and image processing: http://www.csse.uwa.edu.au/~pk/Research/MatlabFns/


Line 58: Line 59:


==Digital communications in medicine (DICOM)==
==Digital communications in medicine (DICOM)==
At an early development stage. [http://octave.svn.sourceforge.net/viewvc/octave/trunk/octave-forge/extra/dicom/ octave-forge svn]
There is a DICOM package hosted on sourceforge: [[dicom | more information]].
*Maintainer: Andy Buckle [http://sourceforge.net/sendmessage.php?touser=1760416contact]
*Currently worked on by: Kris Thielemans
 
===Supporting library===
[http://sourceforge.net/apps/mediawiki/gdcm/index.php?title=Main_Page GCDM]. I think the libraries must be built. Getting source with subversion.
 
svn co https://gdcm.svn.sourceforge.net/svnroot/gdcm/branches/gdcm-2-0
 
cmake and g++ are required. create a gdcmbin directory parallel to the gdcm-2-0 directory. Change to gdcmbin and then run cmake (Linux/cygwin).
 
> mkdir gdcmbin
> ls
gdcmbin gdcm-2-0
> cd gdcmbin
> cmake -DGDCM_BUILD_SHARED_LIBS=ON ../gdcm-2-0/
 
This creates a Makefile in your gdcmbin directory. You can easily edit cmake options with "make edit_cache", for example to build applications like gdcmdump. "make help" is also very useful.
 
make install
 
On my system, I also had to add this to my ~/.bashrc.
 
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/lib
 
Mingw32 is similar. Get openSSL: [http://www.slproweb.com/products/Win32OpenSSL.html Win32 Openssl v1.0.0a Light]
 
cmake -G "MinGW Makefile" -DGDCM_BUILD_SHARED_LIBS=ON ..\gdcm-2-0
mingw32-make all
 
===Features===
All of these still need work, but they exist, and are usable to some extent.
*dicomdict (load different dict - file follows same format as Matlab's)
*dicominfo (load metadata into octave struct. nasty bug in SQ loading)
*dicomlookup (keyword <-> tag)
*dicomread (load pixel data, 2D or 3D int types. Does not yield colormaps, yet.)
*dicomwrite (only does some 2D int types, ATM. No metadata control, yet)
 
===Todo===
 
*dicomanon
*dicomuid
 
===Questions===
I don't have a Matlab license available. I would like to make this package compatible with the dicom functions in Matlab's image processing toolbox, as far as possible.
 
*Octave (and Matlab) stores images (y,x) and DICOM is intrinsically (x,y). Does Matlab transpose images when it loads them?
**matlab reads the data from the dicom file as if it's a raw block of numbers (and then converts if necessary). Therefore, current Octave/dicom behaviour is compatible with matlab.
*I would like people to try m-files that worked with Matlab to let me know of problems.
*(not necessarily) Matlab related: I need examples of odd DICOM files. I have plenty with complex metadata, but I need some with unusual images.
**More than one sample per pixel
**Colour
**Float types
*DA (date) VR: does Matlab turn them into a string?
*if the user supplies metadata (eg HighBit?, BitsStored?) that describe the type of the pixels and this is not in agreement with the type of the image matrix supplied, does Matlab
**silently convert the image pixel type to match the metadata?
**silently convert the metadata to match the pixel type?
**error and do nothing?
*What does dicominfo do when a tag is not in its dictionary: skip it or give error? I was wondering about turning the tag into a variable name, something like Tag_3243_0010. (Matlab 6.5 (2002): Private__3243_0010)
*dicominfo: Items in sequences are not necessarily the same, so cannot be stored in arrays of structs. (Matlab 6.5 (2002): makes nested structs like dcm.RTDoseROISequence?.Item_1.DoseUnits?)
 
== Fixed point toolbox ==
(initial announcement can be found [http://www.octave.org/octave-lists/archive/help-octave.2004/msg01274.html here], with the corresponding thread)
 
When implementing algorithms in hardware, it is common to reduce the accuracy of the representation of numbers to a smaller number of bits. This allows much lower complexity in the hardware, at the cost of accuracy and potential overflow problems. Such representations are known as fixed point.
 
OctaveForge now contains [http://octave.sourceforge.net/Fixedpoint/index.html a toolbox] to perform such fixed point calculations. This toolbox supplies a fixed point type that allows Octave to model the effects of such a reduction in accuracy of the representation of numbers. The major advantage of this toolbox is that with correctly written Octave scripts, the same code can be used to test both fixed and floating point representations of numbers.
 
What it does is create several new user types for fixed point scalar, complex scalars, matrices and complex matrices, and the corresponding operators on these types. As this code was first written against 2.1.50 there is no capabilities at this time for NDArray operations with this code, however I'm not sure this is a problem.
 
A typical use of the toolbox might be something like
 
n = 2;
a = rand (n, n);
b = rand (n, n);
## Create fixed-point version with 1 bit before decimal and 5 after.
af = fixed (1, 5, a);
bf = fixed (1, 5, b);
c = myfunc (a, b);
cf = myfunc (af, bf);
function y = myfunc (a, b)
    y = a * b;
endfunction
 
where as you can see the underlying function myfunc is unchanged, while it is called with either floating or fixed point types. The case above is for fixed-point values with 1-bit before the decimal point and 5 after, and for me gave a result of
 
octave:8> c
c =
  0.98105  0.94436
  0.82622  0.30831
octave:9> cf
cf =
  0.93750  0.90625
  0.78125  0.25000
 
which clearly shows the loss of precision of a fixed-point algorithm with only 6-bits of precision in a matrix multiply. One gotcha in this toolbox is the use of the concatenation operator "[ ]" which will implicitly reconvert fixed-point values back to floating-point, with Octave 2.1.57 or earlier. This is due to an internal limitation of octave that was removed in Octave 2.1.58. So for the best experience it is suggested you use this toolbox with octave 2.1.58 or later.
 
This package is only available with recent versions of OctaveForge (20040707 or later). With the package installed online help is available with the command
octave:1> fixedpoint info
 
As this package is relatively new, all feedback on its use would be most welcome.
 
Matlab also recently introduced a [http://www.mathworks.com/products/fixed Fixed Point Toolbox]. The Octave toolbox has been written independently of the Matlab toolbox and doesn't follow the same syntax. This might change in the future, if the author (DavidBateman) can be bothered to put the effort into making the required changes. Mathworks [http://www.mathworks.com/access/helpdesk/help/pdf_doc/fixedpoint/FPTUG.pdf documentation] for their fixed point toolbox also includes a generic discussion of fixed point numbers that might be a useful addition to the documentation supplied with Octave itself. Another short introduction to fixed point arithmetics is [http://home.earthlink.net/~yatescr/fp.pdf this].


==Adding a quadl like integrator to octave==
==Adding a quadl like integrator to octave==
Line 185: Line 78:
A set of tools for modelling dynamic physical systems using the bond-graph methodology and transforming these models into representations suitable for analysis, control and simulation. These tools use, and generate m files for, GNU Octave.
A set of tools for modelling dynamic physical systems using the bond-graph methodology and transforming these models into representations suitable for analysis, control and simulation. These tools use, and generate m files for, GNU Octave.


==Optimisation==
==Optimization==
There is a variety of code in the octave-forge optimization package http://octave.sourceforge.net/optim/index.html
There is a variety of code in the octave-forge optimization package http://octave.sourceforge.net/optim/index.html


Line 202: Line 95:
*[http://www.statsci.org/matlab/statbox.html Statbox] has some regression functions which are not in Octave. Don't know license or Octave compatibility.
*[http://www.statsci.org/matlab/statbox.html Statbox] has some regression functions which are not in Octave. Don't know license or Octave compatibility.


*[http://www.dpmi.tu-graz.ac.at/~schloegl/matlab/NaN/ NaN-toolbox] A statistic-toolbox for Octave and Matlab®. It handles data with and without MISSING VALUES. More background information is provided in [http://www.dpmi.tu-graz.ac.at/~schloegl/matlab/NaN/README.TXT README]
*[http://pub.ist.ac.at/~schloegl/matlab/NaN/ NaN-toolbox] A statistic-toolbox for Octave and Matlab®. It handles data with and without MISSING VALUES. More background information is provided in [http://pub.ist.ac.at/~schloegl/matlab/NaN/README.TXT README]
 
*[http://www.cepremap.cnrs.fr/dynare/ Dynare] is a pre-processor and a collection of MATLAB or SCILAB routines which solve non--linear models with forward looking variables. Cannot find any insight about the license terms, but some code is derived work from GPL code, e.g., [https://www.cepremap.cnrs.fr/websvn/filedetails.php?repname=repos+1&path=%2Fdynare_v3%2Fmatlab%2Fkernel_density_estimate.m&rev=6&sc=1]. The code is adapted from Anders Holtsberg's matlab toolbox (stixbox).


===Multivariate kernel density estimation===
===Multivariate kernel density estimation===
Line 219: Line 110:


*[http://ltilib.sourceforge.net/doc/homepage/index.shtml LTI-Lib] is an object oriented library with algorithms and data structures frequently used in image processing and computer vision. LGPL
*[http://ltilib.sourceforge.net/doc/homepage/index.shtml LTI-Lib] is an object oriented library with algorithms and data structures frequently used in image processing and computer vision. LGPL
==Symbolic==
===Symbolic Manipulation===
Octave is a matrix algebra tool, not a symbolic algebra tool, so it cannot perform symbolic operations, such as finding the derivative of sine(x) with respect to x
y = derivative('sin(x)','x'); # Not available in Octave
The below mentioned GiNaC will add symbolic functionality to octave.
===Octave interfaces to GiNaC===
There are two Octave interfaces to GiNaC[1] available
*gino -- http://www.sourceforge.net/projects/mtt (download) provides an Octave interface to GiNaC via ginsh
octave:1> df = g_diff ( "sin(x)" , "x" )
df = cos(x)
octave:2> x = pi; eval ( df );
ans = -1
*OctaveForge provides direct access to GiNaC in the [http://octave.sourceforge.net/index/navsymbolic.html symbolic package]. Here is a http://wiki.octave.org/wiki.pl?OctaveForgeGNCIntro short tutorial] to get you started.
===Numerical methods===
In some instances, Octave can give reasonable results using numerical methods
*Octave supports numerical derivatives:
t = linspace(0,6,100);
y = sin(t);
dydt = diff(y) ./ diff(t)
*OctaveForge has other numerical derivative functions.
===Symbolic algebra systems===
Software other than Octave may be more useful for work involving symbolic manipulations
Free software
*Axiom: http://www.nongnu.org/axiom/
*GiNaC/ginsh: http://www.ginac.de
*Maxima: http://maxima.sourceforge.net
*Yacas: http://yacas.sourceforge.net
Non-free software
*Maple: http://www.maplesoft.com
*Reduce: http://www.uni-koeln.de/REDUCE
Note that most of these systems (and Octave!) may be used from within TeXmacs http://www.texmacs.org
A list of some other scientific software is available at http://sal.kachinatech.com/

Latest revision as of 21:30, 13 February 2018

Chaos and fractal contributions and toolboxes[edit]

Econometrics[edit]

  • Michael Creel's Econometrics notes and octave files: http://pareto.uab.es/mcreel/Econometrics/ (try the Econometrics/MyOctaveFiles directory too)
  • Ludwig Kanzler's Econometric software (much of it adapted to Octave):
  • Dynare is a software platform for handling a wide class of economic models, in particular dynamic stochastic general equilibrium (DSGE) and overlapping generations (OLG) models. Licensed under GNU GPL (v3 or later). Consists of a pre-processor (in C++) and of a collection of Octave/Matlab routines.

Communications Toolbox[edit]

OctaveForge now contains a fairly complete implementation of a Matlab R13 compatible communications toolbox http://octave.sourceforge.net/communications/index.html. This includes

  • Random signal creation/analysis
  • Source coding (Quantization, Compand)
  • Block coding (BCH, Reed-Solomon)
  • Modulations (QASK, PSK, etc)
  • Galois fields of even order [i.e. GF(2^m) ]

With a recent version of the OctaveForge installed (version 20030602 or later), online help is available with the command

octave:1> comms info

There are still functions missing or with limitations, and the documentation is not complete. You are therefore welcome to send patches.

Various Communications-Related Octave Code[edit]

Data Formats[edit]

A collection of various signal data formats can be found here. Many formats are supported by the BIOSIG-toolbox for Octave and Matlab.

Speech and Signal Processing Links[edit]

The following packages contain routines that can be made to work with Octave. Sometimes a little effort is required.

The following used to be available:

Other Software:

Digital communications in medicine (DICOM)[edit]

There is a DICOM package hosted on sourceforge: more information.

Adding a quadl like integrator to octave[edit]

quadl is an adaptive quadrature function available in MATLAB (from version 6 I think). It is very fast and accurate and surprisingly it's a small m-file. The algorithm is based on works by Prof. Walter Gander.

Prof. Gander has two m files in his web page http://www.inf.ethz.ch/personal/gander/ which implement his algorithm along with a paper describing the algorithm.

I have not delved deeply into the subject, but the adaptlobe.m function looks equivalent to quadl. Compatibility with MATLAB could be achieved by writing a small quadl wrapper function which simply calles adaptlobe, or by changing adabtlobe's name to quadl.

I have contacted Prof. Gander to allow for distribution of his m-files in octave-forge.

quadl, from the above source is now incorporated into octave-forge-2004-02-12.

Model transformation tools[edit]

http://mtt.sourceforge.net

A set of tools for modelling dynamic physical systems using the bond-graph methodology and transforming these models into representations suitable for analysis, control and simulation. These tools use, and generate m files for, GNU Octave.

Optimization[edit]

There is a variety of code in the octave-forge optimization package http://octave.sourceforge.net/optim/index.html

Parallelism[edit]

Operating Octave remotely and in parallel

Statistics[edit]

Statistics Functions[edit]

  • [ http://www.maths.lth.se/matstat/stixbox/ Stixbox] supports Octave. It has a variety of functions not available in octave/octave-forge such as bootstrap and jackknife estimators and a variety of qq plots. GPL.
  • Statbox has some regression functions which are not in Octave. Don't know license or Octave compatibility.
  • NaN-toolbox A statistic-toolbox for Octave and Matlab®. It handles data with and without MISSING VALUES. More background information is provided in README

Multivariate kernel density estimation[edit]

  • A google search for epanechnikov at the r-project.org site returns several R packages on this subject. They all should be GPL compatible. [1]
  • Weka is written in Java, again, don't know how difficult... look for a file called LWL.java
  • LTI-Lib is an object oriented library with algorithms and data structures frequently used in image processing and computer vision. LGPL

Symbolic[edit]

Symbolic Manipulation[edit]

Octave is a matrix algebra tool, not a symbolic algebra tool, so it cannot perform symbolic operations, such as finding the derivative of sine(x) with respect to x

y = derivative('sin(x)','x'); # Not available in Octave

The below mentioned GiNaC will add symbolic functionality to octave.

Octave interfaces to GiNaC[edit]

There are two Octave interfaces to GiNaC[1] available

octave:1> df = g_diff ( "sin(x)" , "x" )
df = cos(x)
octave:2> x = pi; eval ( df );
ans = -1

Numerical methods[edit]

In some instances, Octave can give reasonable results using numerical methods

  • Octave supports numerical derivatives:
t = linspace(0,6,100);
y = sin(t);
dydt = diff(y) ./ diff(t)
  • OctaveForge has other numerical derivative functions.

Symbolic algebra systems[edit]

Software other than Octave may be more useful for work involving symbolic manipulations

Free software

Non-free software

Note that most of these systems (and Octave!) may be used from within TeXmacs http://www.texmacs.org

A list of some other scientific software is available at http://sal.kachinatech.com/