Level 2 Help for STATS

INP

STRING - Input data sets used by stats.  If input data is in MSS format
then only one data set is required.  Otherwise there must be one data set
per spectral band examined.


OUT

STRING - Output data sets used by STATS.  First data set will contain the
statistics output.  The second output data set (optional) will contain
the input picture with training areas scribed.  (If scribing is requested
and their is only one output file, the output file is assumed to be the
scribed image, and no statistics output file is generated.)

The statistics output file is a classification statistics file (statistics
data set).  Each record (excluding labels) in a classification statistics file
contains statistics for a particular class of pixels.  The statistics are
derived from the multispectral (multi-channel) data.  The size of records in
the file is based on the number of channels (bands).  If 'N' is used to denote
the number of channels, then each record contains the following data in the
order shown. 

   *  2 words (8 bytes) of class name in character (ASCII) format.  This
      takes the form 'CLASSnnn' where nnn is the class number.
      (nnn is left-justified with trailing blanks.)

   *  N words for mean values for this class for each of the N channels.
      Each mean value is stored in REAL*4 format.

   *  1 word, stored in INTEGER*4 format, containing the number of pixels 
      in this class.

   *  N*(N+1)/2 words containing the variances and covariances for the N
      channels for this class.  Each variance and covariance is stored in
      REAL*4 format.

The VICAR system label contains the number of records in the number of lines
field and the number of bytes per record in the number of samples field.


SIZE

INTEGER - Standard Vicar size field (Not used - can be defaulted).


MSS

INTEGER - Denotes that input data is in MSS format and contains the
supplied number of spectral bands.


BAND

INTEGER - Specifies which MSS format bands are to be used as input (The
default is for all bands specified by the MSS parameter).
Syntax is:   BAND = (b1,...)   where b1,... is a list of bands.


DN

INTEGER - Denotes that the outline will be the given DN in scribing the
training areas.  The default is to use 0 or 255 depending on which contrasts
better with the image.


SCRIBE

the training classes scribed on the image.


HIST

INTEGER - Denotes that histograms of the training classes from the
specified bands are to be printed on the line printer.  If the value of
0 is used, histograms for all bands are printed.
Syntax is:   HIST = (b1,...)   where b1,... is a list of bands.

The histogram output uses vertical bars to show the number of pixels in the
DN bins.  The bars are made of a column of  * symbols with a number (or other
symbol) on top.  The number of pixels represented by each * is shown at the
top of the histogram.  The number on top of a column is an additional number of
pixels in the bin beyond those represented by *s.  The letters A through Z are
used to represent the values 10 through 35, respectively.  The $ symbol
represents a value greater than 35. 


SPEC

INTEGER - Denotes that a spectral plot of training classes from the given
bands is to be printed.  If the value of 0 is used, all the bands are 
included.
Syntax is:   SPEC = (b1,...)   where b1,... is a list of bands.


SPLOT

INTEGER - Limits spectral plots to the specified classes (default is that
all classes are included).
Syntax is:   SPLOT = (c1,...)   where c1,... is a list of classes.


EXCLUDE

INTEGER - Ignore any pixel which has the same DN value as one specified.
There is a maximum of 10 DN values that can be specified.
Syntax is:   EXCLUDE = (d1,...)   where d1,... is a list of DNs.


NOPRINT

KEYWORD - Controls the printing of means, standard deviations, covariance
matrices, and training area coordinates on the line printer (default is
to PRINT this information).


VERT

INTEGER - Specifies which  training areas defined by the CLASSx keyword
are to be read in the vertices format.
 The training areas for the classes may be specified either in rectangular
or vertices format.  stats no longer allows some training fields in a given
training area to be in vertices format and some to be in rectangular
format.  Rectangular format is the default and is thus not 
specified explicitly.   The RECT parameter (which was present in the IBM
version of stats), for expliciltly specifying rectangular format, is no
longer used.  The VERT parameter is used to list any and every class using the
vertices format. 


CLASS

Actual keyword is of the form CLASSx where x is a number from 1 to 50.
The number defines which class number this training set belongs.
e.g.  CLASS69 SL,SS,NL,NS    or    CLASS32  L1,S1,L2,S2,...
See CLASSx for a more complete definition.


CLASSX

INTEGER - Denotes the training area information in either rectangular or
vertices formats (default is rectangular).  Rectangular coordinates are
of the form: SL,SS,NL,NS   SL,SS,NL,NS ...  (the standard Vicar size field) 
while vertices format is of the form: L1,S1,L2,S2,... etc.  The range of values
for 'x' is from 1 to 50 such that the valid keywords would be of the form:
CLASS1, ... CLASS50. 
  The training area for the class is composed of one or more training fields.
Each training field is a closed region of the image.  The format must be 
the same for all fields within a class.  In rectangular format the training 
fields are defined by starting line, starting sample, number of lines,
and number of samples.
  In vertices format, only the vertices (line number and sample number) of an
irregular area need be stated.  The program interpolates between each vertex to
determine the perimeter of the training field.  Several rules must be adhered
to in using the vertices format:

1)  The first vertex entered must be the topmost point of the area; that is, 
    the lowest line number.
2)  The vertices must follw in a clockwise order. 
3)  The last vertex must be the same as the first; that is, close the area
    for each training field.

  The maximum size for any training field, whether rectangular or vertices,
is 600 lines by 600 samples.


CLASS1

INTEGER - See CLASSx for a proper definition.


CLASS2

INTEGER - See CLASSx for a proper definition.


CLASS3

INTEGER - See CLASSx for a proper definition.


CLASS4

INTEGER - See CLASSx for a proper definition.


CLASS5

INTEGER - See CLASSx for a proper definition.


CLASS6

INTEGER - See CLASSx for a proper definition.


CLASS7

CINTEGER - See CLASSx for a proper definition.


CLASS8

INTEGER - See CLASSx for a proper definition.


CLASS9

INTEGER - See CLASSx for a proper definition.


CLASS10

INTEGER - See CLASSx for a proper definition.


CLASS11

INTEGER - See CLASSx for a proper definition.


CLASS12

INTEGER - See CLASSx for a proper definition.


CLASS13

INTEGER - See CLASSx for a proper definition.


CLASS14

INTEGER - See CLASSx for a proper definition.


CLASS15

INTEGER - See CLASSx for a proper definition.


CLASS16

INTEGER - See CLASSx for a proper definition.


CLASS17

INTEGER - See CLASSx for a proper definition.


CLASS18

INTEGER - See CLASSx for a proper definition.


CLASS19

INTEGER - See CLASSx for a proper definition.


CLASS20

INTEGER - See CLASSx for a proper definition.


CLASS21

INTEGER - See CLASSx for a proper definition.


CLASS22

INTEGER - See CLASSx for a proper definition.


CLASS23

INTEGER - See CLASSx for a proper definition.


CLASS24

INTEGER - See CLASSx for a proper definition.


CLASS25

INTEGER - See CLASSx for a proper definition.


CLASS26

INTEGER - See CLASSx for a proper definition.


CLASS27

INTEGER - See CLASSx for a proper definition.


CLASS28

INTEGER - See CLASSx for a proper definition.


CLASS29

INTEGER - See CLASSx for a proper definition.


CLASS30

INTEGER - See CLASSx for a proper definition.


CLASS31

INTEGER - See CLASSx for a proper definition.


CLASS32

INTEGER - See CLASSx for a proper definition.


CLASS33

INTEGER - See CLASSx for a proper definition.


CLASS34

INTEGER - See CLASSx for a proper definition.


CLASS35

INTEGER - See CLASSx for a proper definition.


CLASS36

INTEGER - See CLASSx for a proper definition.


CLASS37

INTEGER - See CLASSx for a proper definition.


CLASS38

INTEGER - See CLASSx for a proper definition.


CLASS39

INTEGER - See CLASSx for a proper definition.


CLASS40

INTEGER - See CLASSx for a proper definition.


CLASS41

INTEGER - See CLASSx for a proper definition.


CLASS42

INTEGER - See CLASSx for a proper definition.


CLASS43

INTEGER - See CLASSx for a proper definition.


CLASS44

INTEGER - See CLASSx for a proper definition.


CLASS45

INTEGER - See CLASSx for a proper definition.


CLASS46

INTEGER - See CLASSx for a proper definition.


CLASS47

INTEGER - See CLASSx for a proper definition.


CLASS48

INTEGER - See CLASSx for a proper definition.


CLASS49

INTEGER - See CLASSx for a proper definition.


CLASS50

INTEGER - See CLASSx for a proper definition.