NCD(1)

NAME

ncd - compute the Normalized Compression Distance

SYNOPSIS

ncd  [  -c compressor ] [ -o filename ] [ -bcdhLnqsv ] [-o filestem ] [
-d|f|l|p|t string ] ... [arg1] [arg2]

DESCRIPTION

The Normalized Compression Distance between two objects is defined as

NCD(a,b) = (C(a,b) - min(C(a),C(b))) / max(C(a),C(b))

where

C(a,b) means "the compressed size of the concatenation of a and b"

C(a) means "the compressed size of a"

C(b) means "the compressed size of b"

ncd will print a non-negative number (typically, but not always, 0 <= x < 1.1) representing how different the two objects are. Smaller numbers represent more similar files. The largest number is somewhere near 1. It is not exactly 1 due to imperfections in compression techniques or other irregularities underlying compressor, but for most standard compression algorithms you are unlikely to see a number above 1.1 in any case.

Three compressors are available by default: bzlib, zlib and blocksort. These may be selected with an option in the complearn.conf, see complearn (5) for more details.

ENUMERATION MODES

-f, --file-mode=FILE

select file mode

-l, --literal-mode=STRING

select string literal mode; this is the default. The next argument is a string which, if containing white space, may be enclosed in double-quotes (")

-p, --plainlist-mode=FILE

select plain list mode; argument is a file which contains a list of files to be individually evaluated

-t, --termlist-mode=FILE

select term list mode; argument is a file which contains string literals to be individually evaluated

-d, --directory-mode=DIR

select directory mode; argument is a path which contains files to be individually evaluated

OPTIONS

-c, --compressor=compressor

use and set compressor to use

-L, --list

list available builtin compressors as well as available compression modules. Modules are loaded from the modules subdirectory of /usr/lib/complearn.

-s, --size

get, in place of NCD, the compressed size of a single FILE, STRING, or DIR

-n, --nexus

Nexus output format for distance matrix

-o, --output=FILE

specify binary output filestem, if different from distmatrix, the default. An extension (.clb, .nex, or .txt) will be added, as appropriate to the output file type.

-b, --binary

output results to binary file; the default name is distmatrix.clb

-q, --quiet

suppress ASCII output and messages

-v, --verbose

activate verbose mode

-h, --help

show help options and exit

FILES

$HOME/.complearn/complearn.conf

/usr/share/complearn/complearn.conf

/usr/local/share/complearn/complearn.conf

per-user and system configuration files

see complearn(5) for further details.

$HOME/.complearn/modules

/usr/lib/complearn/modules

standard module automatic loading area. Any shared object compressor

modules found here will be loaded on startup.

ENVIRONMENT

COMPLEARNMODPATH

If this environment variable is set, CompLearn will search the

given directory and load any CompLearn compression modules it finds there (such as the libart.so example included with the CompLearn source distribution) none

DIAGNOSTICS

none

SEE ALSO

anycompress(1), anydecompress(1), complearn(5), maketree(1)

NCD(1)