NAME
add, subtract, multiply, divide, etc.
uspmath - Perform unary math operations on many USP inputs...
SYNOPSIS
uspmath -N ntap -N ntap -N ntap -F ftap -O otap -mode mode
-S scalar -S scalar -S scalar
-B bias -B bias -B bias
-eps eps -min min -max max -old old -new new
-emask emask -nomask -V
DESCRIPTION
uspmath performs unary math operations on multiple input
datasets, creating a single output. Input files must be con-
formable. uspmath is aware of an embedded mask (-emask) in
the data and only performs operations on live samples.
Operations available for multiple inputs are
sum
mult
div
min
max
union
Operations available for a single input file are
clip
change
scale (with bias)
During execution, data scalars and biases are applied before
the requested operation, and inputs are processed in the
order they apear on the command line. Where samples are live
in some inputs and masked in others, the operation is per-
formed on only the live samples.
Using uspmath on the command line, it may easier to list
many input file names in a text file and use -F ftap to read
the list of input names.
ARGUMENTS AND OPTIONS
-N ntap [default: ntap=stdin:]
Enter an input data set name immediately after typing
-N. Repeat as many times as needed for multiple inputs
(see EXAMPLES). Use the name 'stdin:' if the first of
many inputs is a pipe. For only one piped input, -N may
be omitted. File names should include the relative
paths if the files reside in different directories.
-F ftap [default: none]
Enter the name of a file with input file names (one per
line). Scalar and bias values for each file can also be
included, but are optional. The file should look like
this:
indata_1.usp -S1 -B0
indata_2.usp -S2.5 -B0
...
indata_n.usp
If the first file is a pipe, the first name should be
stdin:
-O otap [default: otap=stdout:]
Enter the name for the output volume. This name should
include the relative path if the file is to be placed
in a different directory.
-mode mode [default: mode=sum]
Enter the mode of operation. (each operation follows
scaling)
For multiple inputs, select
sum - sum inputs
mult - multiply inputs
div - divide inputs
min - take minimum of inputs
max - take maximum of inputs
union - create union of live samples
For a single input, select
clip - clip live samples to range
-min:-max
change - change one value to another
-old --> -new
scale - apply scalar and bias only
-S scalar [default: scalar=1.0]
The scalar to apply to an input dataset. Repeat as many
times as needed for multiple inputs (see EXAMPLES).
Scalar input should be ordered to match the input
dataset that the scalar is applied to.
-B bias [default: bias=0.0]
The bias to apply to an input dataset. Repeat as many
times as needed for multiple inputs (see EXAMPLES).
Bias input should be ordered to match the input dataset
that the bias is applied to.
Scalar, bias and operation are applied to input data as
follows:
output = operation( input(1)*scalar(1) + bias(1),
input(2)*scalar(2) + bias(2),
.
.
.
input(n)*scalar(n) + bias(n) )
-eps eps [default: eps=0.0]
-eps is added to all inputs when -mode is 'div'. It is
ignored for all other modes.
-min min, -max max [default: min=-1.0e+37, max=+1.0e+37]
When -mode is 'clip', live samples less than -min are
set to -min and live samples greater than -max are set
to -max.
-old old, -new new [default: old=-1.0e+37, new=-1.0e+37]
When -mode is 'change', samples with the value -old
will be changed to have the value -new. In this mode,
the embedded mask is used during scaling and bias
application but not during the change operation.
-emask emask [default: emask=-1.0e+37]
An embedded mask. Samples with this value do not
undergo scaling, bias or operation.
-nomask [default: .FALSE.]
Use -nomask speed operation when there is no embedded
mask in the data.
-V [default: .FALSE.]
Enter -V to enable verbose printout.
Enter the command line argument -h to get online help. The
program terminates after the help screen is printed.
EXAMPLES
uspmath -Nin1.usp -Nin2.usp -Nin3.usp \
-S0.5 -S1.0 -S0.5 -Oout.usp \
-mode sum -nomask -V
This will add 3 datasets together with weights 0.5, 1.0, and
0.5. uspmath will ignore any embedded mask if present.
uspmath -Nstdin: -Nin2.usp -Nin3.usp \
-Oout.usp -mode sum -emask -99999 -V
This will add 3 datasets together. The first dataset is on a
pipe. The emmbedded mask -99999 is preserved in the output
uspmath -Nin1.usp -Nin2.usp -Nin3.usp \
-eps0.0001 -Oout.usp \
-mode div -emask -99999 -V
This will perform (in1+eps)/(in2+eps)/(in3+eps) where sample
values are not equal to -99999
uspmath -Nin1.usp -Nin2.usp -Oout.usp \
-mode union -emask 0
This will create a union of live values in the two input
datasets. Where both datasets are live, the value in the
first dataset is taken.
uspmath -Nin.usp -Oout.usp -emask-99999 \
-mode change -old -99999 -new 0
This will change the embedded mask of the input from -99999
to 0.
BUGS
Multiple inputs work on comand line but not within ikp.
AUTHOR
M.J. O'Brien - Allied Geophysics, Inc.
COPYRIGHTS
Copyright 2001, Allied Geophysics, Inc. All Rights
Reserved.
Portions of this application are also protected as follows:
copyright 2001, Amoco Production Company
All Rights Reserved
an affiliate of BP America Inc.
Man(1) output converted with
man2html