topic Re: Box and whisker plot generation in Graphics Programming

Box and whisker plot generation

ArchanaSudhir — Mon, 02 May 2016 11:08:05 GMT

I have a data&colon;

Data have;

Input concentration;

49 50 53 56 59 60 40 30 44

How do I generate a box and whisker plot for this with 25th and 75th percentile?

THnaks

Archana

Re: Box and whisker plot generation

BrunoMueller — Mon, 02 May 2016 12:33:47 GMT

Please have a look at the excellent blogs by @Jay54 http://blogs.sas.com/content/graphicallyspeaking/?s=box+and+whisker

Bruno

Re: Box and whisker plot generation

DanH_sas — Mon, 02 May 2016 13:23:10 GMT

Here you go:

Data have;
Input concentration;
cards;
49
50
53
56
60
40
30
44
;
run;

proc sgplot data=have;
hbox concentration;
run;

Re: Box and whisker plot generation

ArchanaSudhir — Mon, 02 May 2016 13:41:26 GMT

Thanks, what does the centre line in hte box mean? Is it not the median?

Thanks

Archana

Re: Box and whisker plot generation

DanH_sas — Mon, 02 May 2016 13:42:44 GMT

Yes, the line is the median, and the marker is the mean.

Re: Box and whisker plot generation

ArchanaSudhir — Mon, 02 May 2016 17:30:49 GMT

Thanks a lot DanH_sas!

One last question, does mean needs to be plotted or geometric mean?

Thanks for all the help, appreciate it:)

Re: Box and whisker plot generation

DanH_sas — Mon, 02 May 2016 17:38:56 GMT

The mean on the box plot is the arithmetic mean, not the geometric mean.

Re: Box and whisker plot generation

Rick_SAS — Mon, 02 May 2016 17:47:14 GMT

The following link is to SAS documentation that you can use to learn more about what each part of a box plot represents:

Schematic box-and-whisker plot.

The SGPLOT procedure varies a little (using a diamond instead of a cross for the mean), but the basic idea applies to the VBOX statement in PROC SGPLOT.

Re: Box and whisker plot generation

ArchanaSudhir — Tue, 03 May 2016 14:47:30 GMT

HI,

1. I am using SAS 9.2. And the Box plots are not getting generated with the code mentioned.

Results has plot. But when I open it just says loading and never loads.

2. What does " whiskers to the last data point within 1.25 times the interquartile range" mean?

Thanks,

Archana

Re: Box and whisker plot generation

Rick_SAS — Tue, 03 May 2016 14:55:18 GMT

I strongly encourage you to read about box plots and the statistics that they represent. Here is a step-by-step introduction:

http://bolt.mph.ufl.edu/6050-6052/unit-1/one-quantitative-variable-introduction/boxplot/

The interquartile range is the value IQR = P75 - P25 where P75 is the 75th percentile value and P25 is the 25th percentile value.

The IQR is used to determine the length of the whiskers. Some software uses 1.25 as a multiplier, others use 1.5.