SAS for R Users

6Frequency Distributions and Cross Tabulations

We have used numeric summary methods in the previous chapter to see analysis of numerical variables. In this chapter we learn how to use frequency distributions and cross tabulations to analyze categorical or factor variables.

6.1 Frequency Distributions in SAS

Proc Freq is the primary procedure for frequency analysis.

proc freq data=sashelp.cars;
   tables origin;
 run;

The FREQ Procedure

Origin	Frequency	Percent	Cumulative Frequency	Cumulative Percent
Asia	158	36.92	158	36.92
Europe	123	28.74	281	65.65
USA	147	34.35	428	100.00

We can use nlevels to find the number of levels in that variable. We can also do frequency analysis on two or more variables by keeping them in tables with a white space between each.

proc freq data=sashelp.cars nlevels;
   tables origin type;
 run;

The FREQ Procedure

Number of Variable Levels
Variable	Levels
Origin	3
Type	6

Origin	Frequency	Percent	Cumulative Frequency	Cumulative Percent
Asia	158	36.92	158	36.92
Europe	123	28.74	281	65.65
USA	147	34.35	428	100.00

Type	Frequency	Percent	Cumulative Frequency	Cumulative Percent
Hybrid	3	0.70	3	0.70
SUV	60	14.02	63	14.72
Sedan	262	61.21	325	75.93
Sports	49	11.45	374	87.38
Truck	24	5.61	398	92.99
Wagon	30	7.01	428	100.00

We can remove the cumulative frequency, percent, row, col., cumulatives by using the following respectively (norow, nocol,nopercent,nocum).

proc freq data=sashelp.cars ...