*
* NEXT ANALYSIS
* *************
*
* This is the Parameter settings file of  MDS by distance smoothing.
* Run the program by typing 'smooth>smooth.out' from the command line.
* Whether or not smoothing is used is controled by the parameter Neps,
* the number of smoothing steps.
* A line that starts with a * is a comment line.                             
* The order of the lines is strict, and no setting can be ommited.           
* The lines cannot be wider than 80 characters                               
*                                                                            
* Description of the first input line: 
*
* Nobs   = 0  listwise data
*        > 0  number of observations
* Dim    > 0  number of dimensions
* q      > 1  Minkowski parameter
*        = 1  city block distances
*        = 2  Euclidean distances
*        > 20 Dominance distance
* MaxIter >0  Maximum number of iterations
* Conv   > 0  Convergence criterion: iteration stops if two subsequent Stress values
*             are smaller than this value
* Trans  > 0  Transformation of dissimilarities
*        = 1  ordinal, primary approach to ties                              
*          2  ordinal, secondary approach to ties                            
*          3  nominal                                                        
*          4  interval                                  
*          5  absolute (no optimal scaling)                                  
*          6  ispline  (knots and order specified on second card)            
*          7  spline + ordinal, primary approach to ties                              
*          8  spline + ordinal, secondary approach to ties   
* NRan   > 0  Number of random starts
* SeedRandom = value used as a seed for the random generator
* Similarity = 0  Data are dissimilarities
*            = 1  Data are similarities
*                                                                   
* Nobs  Dim  q  MaxIter  Conv  Trans   NRan SeedRandom  Similarity
  13    2    2   1000   1.D-8     5     10   315747       0    
*                                                                            
*
* Specification of smoothing options and algorithm settings
* Type   = 1  Use majorization algorithm (recommended)
*        = 2  Use Spectral Gradient Algorithm (can be fast, but no convergence guaranteed).
* Eps    = size of the initial smoothing epsilon used by distanc smoothing
* Neps   = number of smoothing steps (20 is usually a good idea)
* Relax  = use majorization acceleration (speeds up the convergence rate by factor 2)
* Init   = 1 inital configuration is classical scaling                       
*          2 random configuration                                            
*          3 read initial configuration   
* Read weights = 1  weights (per dissimilarity) are specified in the data file 
*              = 0  no weights are read
* Read labels  = 1  object labels are specified in the data file.
*              = 0  no object labels are used
* Optimal dilation  = 1  Optimal dilation for coordinates is computed in each iteration
*                   = 0  No optimal dilation 
* Knots   = k >0   Number of interior knots needed when using spline transformations
* Order   = r >1   Order of the spline transformation
* MaxIn   = k >0   Maximum number of inner iterations using ANNLS
* MaxOut  = k >0   Maximum number of outerer iterations when using ANNLS
*
*
*                           Read    Read   Opt.                
*  Type Eps Neps Relax Init weights labels dil  Knots Order MaxIn MaxOut
     1   0   20    1     1    0       0     1     4     2    0      0 
* 
* Print options (k=0: Nothing printed)
* RawDis   = k  Print k rows of raw dissimilarities 
* StanDis  = k  Print k rows of standardized dissimilarities
* Weights  = k  Print k rows of weights
* InitX    = k  Print k rows of intial coordinates
* FinX     = k  Print k rows of final coordinates
* Dispar   = k  Print k rows of final disparities
* Interm   = 1  Print summary results of smoothing steps
* PrRan    = 1  Print results for random starts
* Hist     = 1  Print history of iterations (also of smoothing steps, much output!).
* Kyst input = 1 Produce file that can be used for Kyst input
* 
*                                                             Print   
* Print no. of rows of:                                       Kyst
* RawDis StanDis Weights InitX FinX  Dispar Interm PrRan Hist Input   
    10      10       0     0    10    0       1      0    0     0
*                                                                            
* Plot options for simple line plots
* FinalConf = 1   Plot final configuration
* Shepard   = 1   Plot Shepard diagram
* Transform = 1   Plot transformation
* Lpi       > 0   Lines per inch (used to produce square line plots)
* Cpi       > 0   Characters per inch (used to produce square line plots)
* LineWidth > 0   Number of characters per line (only used for the plots)
* PageLength> 0   Number of lines per page (only used for the plots)
* 
* FinalConf  Shepard  Transform   Lpi  Cpi  LineWidth  PageLength  
     1          1       1          20   38    80           24        
*
* Name of the data file
funk.dat
*
* Fortran format of the dissimilarities. '(*)' is Free format
(*)
*
* Fortran format of the weights.'(*)' is Free format
(*)
*
* Fortran format of the initial configuration. '(*)' is Free format
(*)
* Name of the data file that can be used for KYST input
kyst2a.in
