DATAMANIP commands


PROGRAM NAME

kexpon - Introduce Exponential Noise in Input Object

DESCRIPTION

kexpon introduces exponential noise of the specified parameters in the input data object (-i). The noise can either replace the existing value data or it can be added to the existing value data by specifying the appropriate flag (-replace or -add). The variance (-evar) of the exponential distribution must be specified.

Data Type The data type of the output object is the same as the input object's data type. Internally, the data is processed using one of the following: unsigned byte, long, unsigned long, double, or double complex. Data will not be cast to a lower type for processing.

Map Data If the input object contains a map, then the operation is performed on the map data. If the input object does not have a map but has value data, then the operation is performed on the value data.

Validity Mask If there is a validity mask associated with the input object, the mask is transferred to the output object. To decrease processing time, the operation will still be performed on those data that have corresponding mask values equal to zero. If the data under the mask is genuinely fInotfP processable, such as NaN or infinity, the user can specify that a value be substituted in place of the original value when it is accessed. The masked data substitution attributes for a data object can be explicitly set and stored using the fISet AttributefP operator, DATAMANIP::ksetdattr.

Explicit Location and Time Data If the input object has time data or location data, it is transferred to the output object.

Failure Modes This program fails if the input object lacks both map data and value data.

REQUIRED ARGUMENTS

-i
type: infile
desc: input file
-o
type: outfile
desc: output file

Mutually Exclusive Group; you must specify ONE of:

-replace
type: flag
desc: Replace existing values with noise
OR
-add
type: flag
desc: Add noise to the specified segment

OPTIONAL ARGUMENTS

-evar
type: double
desc: Variance for exponential noise
default: 1
bounds: value > 0.0

EXAMPLES

SEE ALSO

DATAMANIP::knoise

RESTRICTIONS

REFERENCES

COPYRIGHT

Copyright (C) 1993 - 1997, Khoral Research, Inc. ("KRI") All rights reserved.