Jean, thanks for your answer. However, I fear gnumeric does something else. I have a data set with one very strong outlier, much farther out than 3* the IQ and according to this definition it should not be drawn. But it does, so it definitely includes the maximum value. I am quite happy gnumeric does not clip the values at 3*IQ, but I still would like to know what it does exactly ... Hannes p.s. here are my data: HCI TT 2010 1 4 4 6 7 8 8 8 11 11 12 13 14 14 26 28 46 598 -----Ursprüngliche Nachricht----- Von: Jean Brefort [mailto:jean brefort normalesup org] Gesendet: Montag, 24. Oktober 2011 10:42 An: Kuehtreiber Hannes Cc: gnumeric-list gnome org Betreff: Re: outlieres in boxplots The answer to your question is at http://itl.nist.gov/div898/handbook/eda/section3/boxplot.htm Regards, Jean Le lundi 24 octobre 2011 à 08:23 +0200, Kuehtreiber Hannes a écrit :
Hello everybody, I am new to gnumeric, yet quite impressed by its statistical abilities. Just did some boxplots, but I am missing information on how gnumeric decides which data are outliers. Could someone enlighten me? Hannes p.s. Tried google, but it did not come up with anything useful. -- Mag. Hannes Kühtreiber Tiroler Landesmuseen-Betriebsgesellschaft m.b.H. Naturwissenschaftliche Sammlungen Feldstraße 11a A-6020 Innsbruck Tel.: +43 (0) 512 59489-425 h kuehtreiber tiroler-landesmuseum at _______________________________________________ gnumeric-list mailing list gnumeric-list gnome org http://mail.gnome.org/mailman/listinfo/gnumeric-list
Attachment:
outlier.jpg
Description: outlier.jpg
