XploRe Help : grboxmediancl

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:	graphic
See also:	grbox grdotcl grhistcl plotboxcl dispboxcl grboxcl grboxmeancl grboxmedian

Quantlet:		grboxmediancl
Description:		generates a boxplot of classified data with the median line. The box borders are the percentiles which are equivalent to the mean plus/minus one standard deviation in the normal case (16% and 84% percentile) and the whiskers are equivalent to the mean plus/minus two standard deviations in the normal case (2.5% and 97.5% percentile).

Usage:	`bp = grboxmediancl(gr, fr{, col})`
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
	`col`	(optional), color of the boxplot
Output:
	`bp`	composed graphical object

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("graphic") library("stats") randomize(666) ; make 15 random intervals gr = sort(normal(5)|5+normal(5)|10+normal(5),0) fr = ceil(20*uniform(14)) bp = grboxmediancl(gr,fr,2) d = createdisplay(1,1) show(d, 1, 1, bp)

Result:

Shows a green "robust" boxplot.

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