INPut data sets in byte, I*2, I*4, or REAL*4 format. The format of the input files is taken from its label. There may be 0 to 18 input data sets.
OUTput data set in byte, I*2, I*4, or REAL*4 format. Primary input format is assumed if FORMAT is not specified
(Starting Line, Starting Sample, Number of Lines, Number of Samples) The standard vicar size field. SIZE gives the area of operation on inputs and defines the size of the output. If larger than any of the inputs, its values are reduced to the minimum line or sample of the inputs. The SIZE parameter takes precedence over SS, SL, NS, NL. If it is not given (default case), then it gets the information from the first input file label.
The bands parameter determines the bands in the input files to be processed. It is specified as (SB,NB), where SB is the starting band NB is the number of bands to be copied (Default is all bands in image.)
Starting Line - Alternate method of specifying the first SIZE parameter value
Starting Sample - Alternate method of specifying the first SIZE param value
Number of Lines - Alternate method of specifying the first SIZE param value
Number of Samples - Alternate method of specifying the first SIZE param value
Starting Band of window - Can not be defined in SIZE parameter. (Default is first band in image.)
Number of Bands of window - Can not be defined in SIZE parameter. (Default is number of bands in image.)
Image organization for output image when zero input images. If omitted, output image will be BSQ. Has no effect if there ARE input images.
Valid values: BYTE, HALF, FULL, REAL. These values correspond to the following data formats: byte (BYTE), integer*2 (HALF), integer*4 (FULL), real*4 (REAL). This parameter specifies the data format of the output file(s). Default is primary input format.
Specifies a function to be applied to the input data set. The function, in the form of a FORTRAN or C like expression, to be applied to the input data set(s) to produce the output data set. The default expression is "IN1". The function string can include constants, including constants expressed in floating point notation. The string can also include TCL variable references.
Specifies that the results of the calculation will be truncated. Default is that results will be rounded. This keyword is inoperative if real*4 output is specified.
Array of values to be excluded from function evaluation. Up to twenty (20) values can be specified. If any one of the specified values are encountered in the inputs, that pixel is not evaluated by the function and the replace- ment value specified by the REPLACE parameter is placed in the output pixel. If the REPLACE parameter is not specified, the first excluded value among the input pixels is used as the output pixel.
This is the user defined range of values to be included in function evaluation. If values fall outside of this range, the REPLACE parameter value is placed on the output pixel. If REPLACE is not specified, the lower limit of the range is used. If the REPLACE value is outside the valid data range, e.g. -54000 for halfword data, then the closest valid data value is used in the replacement. If the limits are beyond the valid range of the input data, the extremes of the input data values are used, e.g. LIMITS=(-5,240) for byte data yields true limits of (0,240).
Value to be placed in output file when input value is an excluded value or is outside the user specified limits. REPLACE's value must be within the valid data range of the input file. If the input file is byte, REPLACE can be any value between (0,255), inclusive. If a REPLACE is outside this range, the closest valid data value is chosen, e.g. REPLACE=-40000 for halfword data yields a REPLACE=-32768.
Activating the keyword 'DUMP tells F2 to display the symbolic output of
the compiled code produced by the expression-compiler subroutine KNUTH.
Its purpose is primarily diagnostic, and allows the examination of the
way that KNUTH has interpreted the user's input function.
For example, the command
f2 INP=(F1,F2) OUT=X FUNCTION="IN1+SQRT(IN2)" 'DUMP
will output something like:
Beginning VICAR task F2
SQRT 2
ADD 1
RETN 0
...
Which reads: load the SQRT of register 2 (IN2) into the active
register (which is index 0), then ADD the value of register 1 (IN1)
to it and RETurnN the value in the active register 0 -- which is
indeed the requested operation.
Formerly, F2 always displayed this code dump, but this was sometimes
inconvienient with long, complex expressions whose dump might require
80 or more output lines in a log file.