Source code

• Imlib A small C++ image processing toolbox (source code in C++). Makefiles are included for Linux/Cygwin and Visual Studio 6.0 project files for Win32.

• A spectral fitting package (source code in C++) for Linux/Cygwin. Uses unix sed/awk and Gnuplot as visual feedback.

• "Mathutil" contains various mathematical functions in C++ that I have written myself and copied from others. Examples include cylindrical and spherical Bessel functions (J,Y,K,I), a random number generator with period 2**144 and an approximation of Gaussian probability phi(x) and many other functions. Credits to authors and URL:s are given in the code. In many cases the URL:s in the code don't work but I neverteheless hope it will clear who the real author is.
  
  mathutil.h
  mathutil.cpp
  
  The program "rempol" fits an n:th order polynomial and removes it from the data. The method used is a discrete orthogonal polynomial Fourier (OPF) expansion (which diagonalizes the normal equations) for numerical stability. The program is useful for example when removing wiggly baselines in signals (strutil are helper methods for simple string handling). The OPF expansion gives much more stable results, and large polynomial orders may be used without loosing numerical stability (be careful about overfitting).
  
  strutil.h
  strutil.cpp
  polyfit.h
  polyfit.cpp
  rempol.cpp
  
  Sometimes data smoothing is necessary while retaining the global max-and-min structure of data, such as in the treatment of spectra. The Savitzky-Golay filter satisfies these critera and is based on polynomial interpolation expressed as a FIR-filter. The programs use the newmat11library, which you can either download from the kalman11.zip distribution below or copy from Robert's site to directory next the "src" directory.
  
  savgolayfilter.zip