Dear All,

A first version of the stacking and modelling software is in the /RAID6/insar_lab/matlab directory.  It makes stacks, and models our roi_pac interferograms plus GPS data.  Some example plots
are attached below.  Additional examples are in

file:///RAID6/insar_lab/matlab/examples_MakeModelData/html/example_ML1.html
file:///RAID6/insar_lab/matlab/examples_MakeModelData/html/example_ML2.html
file:///RAID6/insar_lab/matlab/examples_MakeModelData/html/example_Kil1.html
file:///RAID6/insar_lab/matlab/examples_MakeModelData/html/example_Kil2.html

To run it,
 copy setmlibs.m  and the  example_MakeModelData/*min files in your directory and run

setmlibs
MakeModelData(' example_Kil1.min')

The html files are currently produced using
pMakeModelData(*), but this needs some more work

You need matlab R2006a. 

This supposed  to be the first piece of a complex InSAR processing suite. Everything is controlled by an text file (*min). The hope is that everybody in the group will contribute code. If you
have new code (individual files) just put it in your own directory (e.g. ~ngourmelen/matlab/utillib) and modify your setmlibs so that your own code is at the beginning of the path. Once we have
verified that the new code works using the example files we can merge it with the code in the insar_lab directory.  We need a system to control code changes such as CVS.

I expect to continue working on MakeSimpleModel.m. Noel will merge his existing code with this and eventually incorporate our SBAS code (Noel wrote the graphics part.).

regards

Falk

RELEASE_NOTES
- Currently bounds are given in the *min file in the local (disloc) coordinate system. This is not good. Everything
  should be given in geographic coordinates (lat/long) and then transformed into local systems for processing without the user feeling it.
  For NOEL this means that for incorporating Coulomb all we need is a function to transform the input and output files from geogramphic to the COulmb coordinate
  Display of the results should be done in geographic coordinates with PlotIgram,

-The programming style is supposed to be similar to roi_pac in perl. In roi_pac we use
       $restart and cleandir.pl .... 
 and clearndir.pl is only run if $restart exists. 
 In matlab we try to use e.g.
      if Faults   PlotIgram(Faults); end
 The defaults of the variable Faults is 'false' or 0 (logical) and no faults are plotted. However, 
 if the variable Faults contains data 'if Faults' returns 'true' and the faults are plotted. This was the plan 
 when I started. Unfortunatelty this does not always work. When the variable faults contains a NaN we get an error. At some places
 I used if Faults(1) but this does not work either if Faults(1)=0 or Faults(1)=NaN.
 Sometimes there is a need to check for fields in a structure or cells in a cellarray what causes problems. E.g. if Faults.xy gives an error if Faults.xy is not 
 defined. What we probably need is a function
    isdefined()     or
    isnotemptynomatterwhat()
 which first checks whether a variable exist, what type pf variable it is and then whether it contains something other than 'false'.
  Once we have this function it should be used
 throughout the entire code. An example of how ugly simple inquiries can become is
    modellib/MakeBounds.m:if ~(~islogical(disloc_bounds) | disloc_bounds(1) ) && N_disloc
 I wonder why matlab does not come with such a function. Please advice.