A very powerful tool in SAS is the custom format. You can create ranges that will be used by Reporting or Analysis procedures to make groups. Most of the graphing procedures will honor them as well.
Here is an example of creating a data set with data step code (which the preferred way to show data as we can test code against such a data set when we run it). Then create a custom format and demonstrate how Proc Freq will generate the counts, running total, percent of values and running percent.
data have;
input minutes Uv;
datalines;
1 0.229528
2 0.262015
3 0.06782
4 1.263749
5 2.969982
6 2.831766
7 1.749711
8 0.164038
9 0.685758
10 0.924174
11 0.150946
12 0.384327
13 1.135237
14 1.9148
15 1.039687
;
proc format;
value uvrange
low -<1 = 'uV<1'
1 - 2 = 'uV 1 to 2'
2<-high = 'uV>2'
;
proc freq data=have;
tables uv;
format uv uvrange.;
run;
data want;
set have;
Low = (uv < 1);
Mid = (1 le uv le 2)l
high
One of the very powerful advantages of a format is that when the boss asks "what about interval boundaries at 1.5 and 2.5?" or anything similar. The only requirement would the create a new format and use that with the Prod Freq code instead of this Uvrange format.
Note that this works for just about anything where groups ONLY depend on values of a single variable.
If your data had another variable, such as the machine or operator that made the recording you could include that variable on the tables statement to get the counts and percentages for each machine or operator (or test sequence or other identifier like this:
proc freq data=have;
tables uv * operatorid;
format uv uvrange.;
run;
Click image to register for webinar
