Hi, the boxplot is based on the quartiles which are much less outlier sensitive than mean and SD and should therefore not be "heavily distorted by outliers". What you mean is presumably that you see the area of the main bulk of the data only as a very small box on the screen because of your outliers. However, a simple straight forward method for outlier identification is median +/- 5.2*mad as suggested by Hampel, Technometrics 27 (1985) 95-107. Outlier identification by use of mean and SD is often bad because these statistics are strongly influenced by the outliers.
x <- data vector medx <- median(x) madx <- mad(x) outliers <- (x<medx-5.2*madx) | (x>medx+5.2*madx) selected <- x[!outliers] Best, Christian On 20 Feb 2003, Rado Bonk wrote: > Dear R-users, > > I have two outliers related questions. > > I. > I have a vector consisting of 69 values. > > mean = 0.00086 > SD = 0.02152 > > The shape of EDA graphics (boxplots, density plots) is heavily distorted > due to outliers. How to define the interval for outliers exception? Is > <2SD - mean + 2SD> interval a correct approach? > > Or should I define 95% (or 99%) limit of agreement for data interval, > and exclude lower, and higher values? > > II. > How to extract only those values from vector which fulfill the condition > of interval (higher than A, and lower than B)? > > Rado Bonk > > ______________________________________________ > [EMAIL PROTECTED] mailing list > http://www.stat.math.ethz.ch/mailman/listinfo/r-help > -- *********************************************************************** Christian Hennig Seminar fuer Statistik, ETH-Zentrum (LEO), CH-8092 Zuerich (currently) and Fachbereich Mathematik-SPST/ZMS, Universitaet Hamburg [EMAIL PROTECTED], http://stat.ethz.ch/~hennig/ [EMAIL PROTECTED], http://www.math.uni-hamburg.de/home/hennig/ ####################################################################### ich empfehle www.boag.de ______________________________________________ [EMAIL PROTECTED] mailing list http://www.stat.math.ethz.ch/mailman/listinfo/r-help
