Provided by: numdiff_5.9.0-1_amd64 
NAME
numdiff - compare similar files with numeric fields
DESCRIPTION
Usage: numdiff -h|--help|-v|--version or
numdiff [-s IFS][-D DELIMS][-a THRVAL[:RANGE|:RANGE1:RANGE2]][-r THRVAL[:RANGE|:RANGE1:RANGE2]][-2][-F
NUM][-# NUM][-P][-N][-I][-c CURRNAME][-d C1C2][-t C1C2][-g N1N2][-p C1C2][-n C1C2][-e C1C2][-i C1C2][-X
1:RANGE][-X 2:RANGE][-E][-U][-b][-V][-O[NUM]][--raw][-q][-S][-z 1:RANGE][-z 2:RANGE][-Z 1:RANGE][-Z
2:RANGE][-m][-H][-f[NUM]][-T][-B][-l PATH][-o PATH] FILE1 FILE2
Compare putatively similar files line by line and field by field, ignoring small numeric differences
or/and different numeric formats.
RANGE, RANGE1 and RANGE2 stay for a positive integer value or for a range of integer values, like 1-, 3-5
or -7. The two arguments after the options are the names of the files to compare. The complete paths of
the files should be given, a directory name is not accepted. The given paths cannot refer to the same
file but one of them can be "-", which refers to stdin.
Exit status: 1 if files differ, 0 if they are equal, -1 (255) in case of error
-s, --separators=IFS
Specify the set of characters to use as delimiters while splitting the input lines into fields
(The default set of delimiters is space, tab and newline). If IFS is prefixed with 1: or 2:, use
the given delimiter set only for the lines from the first or the second file respectively
-D, --delimiters=DELIMS
Specify the set of strings to use as delimiters while splitting the input lines into fields (The
default set of delimiters is space, tab and newline). If DELIMS is prefixed with 1: or 2:, use
the given delimiter set only for the lines from the first or the second file respectively
-a, --absolute-tolerance=THRVAL[:RANGE|:RANGE1:RANGE2]
Set to THRVAL the maximum absolute difference permitted before two numeric fields are regarded as
different (The default value is zero). If a RANGE is given, use the specified threshold only when
comparing fields whose positions lie in RANGE. If both RANGE1 and RANGE2 are given and have the
same length, then use the specified threshold when comparing a field of FILE1 lying in RANGE1 with
the corresponding field of FILE2 in RANGE2
-r, --relative-tolerance=THRVAL[:RANGE|:RANGE1:RANGE2]
Set to THRVAL the maximum relative difference permitted before two numeric fields are regarded as
different (The default value is zero). If a RANGE is given, use the specified threshold only when
comparing fields whose positions lie in RANGE. If both RANGE1 and RANGE2 are given and have the
same length, then use the specified threshold when comparing a field of FILE1 lying in RANGE1 with
the corresponding field of FILE2 in RANGE2
-2, --strict
Consider two numerical values as equal only if both absolute and relative difference do not exceed
the respective tolerance threshold
-F, --formula=NUM
Use the formula indicated by NUM to compute the relative errors. If 'NUM' is 0 use the classic
formula. If 'NUM' is 1 compute the relative errors by considering the values in FILE1 as sample
values. If 'NUM' is 2 compute the relative errors by considering the values in FILE2 as sample
values.
-#, --digits=NUM
Set to NUM the number of digits in the significands used in multiple precision arithmetic
-P, --positive-differences
Ignore all differences due to numeric fields of the second file that are less than the
corresponding numeric fields in the first file
-N, --negative-differences
Ignore all differences due to numeric fields of the second file that are greater than the
corresponding numeric fields in the first file
-I, --ignore-case
Ignore changes in case while doing literal comparisons
-c, --currency=CURRNAME
Set to CURRNAME the currency name for the two files to compare. CURRNAME must be prefixed with 1:
or 2: to specify the currency name only for the first or the second file
-d, --decimal-point=C1C2
Specify the characters representing the decimal point in the two files to compare
-t, --thousands-separator=C1C2
Specify the characters representing the thousands separator in the two files to compare
-g, --group-length=N1N2
Specify the number of digits forming each group of thousands in the two files to compare
-p, --plus-prefix=C1C2
Specify the (optional) prefixes for positive values used in the two files to compare
-n, --minus-prefix=C1C2
Specify the prefixes for negative values used in the two files to compare
-e, --exponent-letter=C1C2
Specify the exponent letters used in the two files to compare
-i, --imaginary-unit=C1C2
Specify the characters representing the imaginary unit in the two files to compare
-X, --exclude=1:RANGE
Select the fields of the first file that have to be ignored
-X, --exclude=2:RANGE
Select the fields of the second file that have to be ignored
-E, --essential
While printing the differences between the two compared files show only the numerical ones
-U, --dummy
While printing the differences between the two compared files neglect all the numerical ones
(dummy mode)
-b, --brief
Suppress all messages concerning the differences discovered in the structures of the two files
-V, --verbose
For every couple of lines which differ in at least one field print an header to show how these
lines appear in the two compared files
-O, --overview[=NUM]
Display a side by side difference listing of the two files showing which lines are present only in
one file, which lines are present in both files but with one or more differing fields, and which
lines are identical. If 'NUM' is zero or is not specified, output at most 130 columns per line.
If 'NUM' is a positive number, output at most 'NUM' columns per line. If 'NUM' is a negative
number, do not output common lines and display at most -'NUM' columns per line.
--raw
Display the differences between the two compared files in raw format (not very convenient for
humans)
-q, --quiet, --silent
Suppress all the standard output
-S, --statistics
Add some statistics to the standard output
-z, --blur-if-numerical=1:RANGE
Select the fields of the first file that have to be blurred during the synchronization procedure
only if they turn out to be numeric
-z, --blur-if-numerical=2:RANGE
Select the fields of the second file that have to be blurred during the synchronization procedure
only if they turn out to be numeric
-Z, --blur-unconditionally=1:RANGE
Select the fields of the first file that have to be unconditionally blurred during the
synchronization procedure
-Z, --blur-unconditionally=2:RANGE
Select the fields of the second file that have to be unconditionally blurred during the
synchronization procedure
-m, --minimal
During synchronization try hard to find a smaller set of changes
-H, --speed-large-files
During synchronization assume large files and many scattered small changes
-f, --test-filter[=NUM]
Run only the filter and then show the results of its attempt to synchronize the two files. If
'NUM' is zero or is not specified, output at most 130 columns per line. If 'NUM' is a positive
number, output at most 'NUM' columns per line. If 'NUM' is a negative number, do not output
common lines and display at most -'NUM' columns per line.
-T, --expand-tabs
Expand tabs to spaces in output while displaying the results of the synchronization procedure
(meaningful only together with option -O or -f)
-B, --binary
Treat both files as binary files (only meaningful under Doz/Windoz)
-l, --warnings-to=PATH
Redirect warning and error messages from stderr to the indicated file
-o, --output=PATH
Redirect output from stdout to the indicated file
-h, --help
Show help message and predefined settings
-v, --version
Show version number, Copyright, Distribution Terms and NO-Warranty
Default numeric format (for both files to compare):
Currency name = ""
Decimal point = `.'
Thousands separator = `,'
Number of digits in each thousands group = 3
Leading positive sign = `+'
Leading negative sign = `-'
Prefix for decimal exponent = `e'
Symbol used to denote the imaginary unit = `i'
COPYRIGHT
Copyright © 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017 Ivano Primi
<ivprimi@libero.it>
License GPLv3+: GNU GPL version 3 or later, see <http://gnu.org/licenses/gpl.html>.
This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent
permitted by law.
SEE ALSO
The full documentation for numdiff is maintained as a Texinfo manual. If the info and numdiff programs
are properly installed at your site, the command
info numdiff
should give you access to the complete manual.