XploRe Help : meancl

Keywords - Function groups - @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Library:	stats
See also:	mean mediancl varcl

Quantlet:		meancl
Description:		Computes the mean of the columns of classified data

Usage:	`z = meancl(gr, fr)`
Input:
	`gr`	m x n grid matrix, each column must be ordered ascendently
	`fr`	(m-1) x n matrix of frequencies for each interval in each column
Output:
	`z`	1 x n row vector of means for each column

Note:

Each column of the gr matrix represents a grid (must be ordered ascendently) and each corresponding column of fr matrix represents the numbers of observations in the corresponding intervals (given by gr matrix). E.g., for gr=#(1,2,3)~#(10,15,20) and fr=#(7,4)~#(14,16) the first dataset contains 7 observations in the interval [1,2) and 4 observations in the interval [2,3). The same is true for the second column. The representative for each interval is the mean of its boundaries.

Example:

library("stats") randomize(666) ; make four columns each with 15 random intervals gr = sort(normal(5,4)|5+normal(5,4)|10+normal(5,4),0) gr[,3] = 10 + gr[,3] fr = ceil(30*uniform(14)~20*uniform(14,2)~2*uniform(14)) meancl(gr,fr)

Result:

Contents of m
[1,]   4.6065   5.3096    14.82   4.6954

Author: J. Budek, K. Komorad, W. Haerdle, 20010716. license MD*Tech