creating a dummy variable to run a linear regression

[sasuser94]

I am attempting to create a dummy variable for gender. Every time i run my code i get all zeros and i do not have an all male sample.



data dummy4;
set dummy1;
if gen = 'male' then gen = 0;
if gen = 'female' then gen =1;
run;


data dummy1;
 input id gen bedsun;
 datalines;
2	Female	8100
3	Female	86400
4	.        84600
5	Female	82800
6	Female	84600
7	Female	82800
8	Female	82800
9	Male	84600
10	Male	84600
11	Male	81900
12	Male	84600
13	Male	12600
14	Male	85500
15	Female	2700
16	Female	7200
17	Female	81000
18	Female	79200
19	Female	86400
20	Female	83700
21	Female	5400
22	Female	78300
23	Female	86400
24	Male	84600
25	Male	900
26	Female	86400
27	Female	77400
28	Female	81000
29	Female	3600
30	Female	900
31	Female	83700
32	Female	86400
33	Female	75600
34	Female	85500
35	Female	81000
36	Male	3600
37	Female	3600
38	Female	74700
39	Female	85500
40	Male	82800
41	Female	18900
42	Female	86400
43	Female	3600
44	Female	81900
45	Female	10800
46	Male	84600
47	Female	83700
48	Male	84600
49	Male	85500
50	Male	8100
51	Female	83700
52	Female	82800
53	Female	5400
54	Female	3600
55	Female	77400
56	Female	9000
57	Female	84600
run;

[PaigeMiller]

This question comes up a lot, and most of the answers are: "don't create a dummy variable yourself to do this".

Use PROC GLM, which will create the dummy variable for you when you use the CLASS statement, and fit the regression line.

You specific code fails because the values have a capital letter, and so your test in the IF statement also needs a capital letter.

if gen='Male' then dummy=0;
else if gen='Female' then dummy=1;

But really, don't create the dummy variables yourself.