DataMuseum.dk

top - download

    Length: 1389 (0x56d)
    Types: TextFile
    Names: »vincent.1«

└─⟦87ddcff64⟧ Bits:30001253 CPHDIST85 Tape, 1985 Autumn Conference Copenhagen
    └─ ⟦this⟧ »cph85dist/stat/doc/man/vincent.1« 

TextFile

.TH VINCENT 1 "March 5, 1985" "UNIX|STAT 5.0" "UNIX User's Manual"
.SH NAME
vincent \- vincentizing of data from files
.SH SYNOPSIS
.B vincent
n file1 file2 ...
.SH DESCRIPTION
.I Vincent
is useful for looking at trends over time/trials in data.
.I Vincent
takes as arguments a series of files of numbers.
The filename ``-'' can be used to read from the standard input.
For each file,
.I vincent
divides the series of numbers into \fIn\fR equal parts
using linear interpolation and prints the means for each part
(called a vincentile).
For example,
.ce
vincent 5 foo bar
will print the means for each fifth of the files "foo" and "bar."
For each file argument,
.I vincent
prints the number of points read in from that file,
the name of the file, and the means for each of the ``vincentiles''
you have asked for.
.PP
.I Vincent
is useful for comparing series of different lengths.
This was its original use in mathematical psychology
some decades ago.  See the Handbook of Mathematical Psychology,
by, among others, Pat Suppes for a thrilling discussion.
.SH AUTHOR
Gary Perlman
.SH SEE\ ALSO
unixstat(1), ts(1)
.br
The program is named after S. B. Vincent who suggested a 4-tile
version of the procedure.
You can look up the original source in his 1912 paper,
The function of the vibrassae in the behavior of the white rat.
.I "Behavior Monographs."
.SH KEYWORDS
statistics, data analysis