Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site brl-vgr.ARPA
Path: utzoo!watmath!clyde!akgua!sdcsvax!sdcrdcf!hplabs!hao!seismo!brl-tgr!brl-vgr!gwyn
From: gwyn@brl-vgr.ARPA (Doug Gwyn )
Newsgroups: net.crypt
Subject: Re: need frequency tables, or reference
Message-ID: <1409@brl-vgr.ARPA>
Date: Sat, 12-May-84 00:10:06 EDT
Article-I.D.: brl-vgr.1409
Posted: Sat May 12 00:10:06 1984
Date-Received: Wed, 9-May-84 03:25:24 EDT
References: <561@aecom.UUCP>
Organization: Ballistics Research Lab
Lines: 18

There is a book entitled something like "Elementary Cryptanalysis" by
Helen F. Gaines that contains lots of frequency tables and so forth.

It is nice however to have a program around for running lots of files
through to build such tables, since different plaintext domains have
different statistics.  For cryptanalysis work you should have table-
driven analysis programs that among other things derive their model
of plaintext characteristics from such tables, much as the Writer's
Workbench derives its style standards from specified model documents.
It is almost trivial to write such a program, especially in a decent
character-handling language like C.  All you have to do is:
	initialize_array_to_all_zero_counts;
	initialize_total_objects_processed_counter_to_zero;
	for each input character, digram, trigram, etc.:
		increment_corresponding_array_location;
		increment_total_objects_processed_counter;
Then the final array elements, divided by the total_objects counter,
are the relative (character, digram, trigram, etc.) frequencies.