R Loop To get Survey Weighted Counts

[Guest]

input person status group angle weight
1 2 3 30 152
2 3 2 29 176
3 2 6 20 146
4 3 6 30 194
5 3 4 21 123
6 2 2 23 169
7 1 3 29 118
8 3 6 30 165
9 3 5 27 130
10 2 1 22 196
11 1 4 24 147
12 1 6 21 172
13 2 2 26 179
14 1 5 26 142
15 3 6 23 170
16 2 1 22 134
17 3 2 26 138
18 1 2 30 129
19 2 5 29 194
20 3 1 20 129
21 1 5 26 190
22 2 3 23 195
23 1 2 24 105
24 3 3 30 136
25 3 5 23 157
26 1 1 22 188
27 3 4 26 134
28 1 5 29 177
29 2 3 27 125
30 1 3 28 185
31 1 1 29 103
32 2 3 26 175
33 2 3 21 151
34 1 2 26 183
35 2 5 22 135
36 2 2 28 163
37 1 3 28 112
38 1 6 22 102
39 1 2 30 170
40 1 6 30 199
41 2 1 26 161
42 2 6 21 183
43 1 1 30 115
44 2 6 28 121
45 2 2 24 140
46 1 4 30 106
47 1 3 25 139
48 3 3 24 115
49 2 5 24 160
50 2 2 23 185
51 2 1 20 185
52 2 2 21 145
53 3 5 27 148
54 3 5 22 108
55 1 2 23 192
56 3 5 22 153
57 1 2 20 178
58 1 2 21 128
59 1 4 29 118
60 1 6 26 102
61 3 2 20 181
62 1 4 30 197
63 3 2 26 147
64 1 5 26 110
65 2 5 26 132
66 3 2 23 132
67 1 6 30 147
68 2 5 23 188
69 1 6 27 194
70 2 6 28 119
71 3 2 25 108
72 3 6 28 126
73 3 2 28 121
74 2 2 23 129
75 3 3 25 121
76 3 1 23 123
77 2 4 26 107
78 1 1 23 149
79 1 6 26 160
80 1 4 21 160
81 2 4 30 147
82 2 4 21 165
83 3 1 29 109
84 1 1 21 181
85 2 3 27 100
86 1 2 27 172
87 3 3 26 120
88 2 4 26 126
89 2 1 30 163
90 2 3 30 135
91 1 6 22 196
92 1 1 28 187
93 3 2 29 169
94 2 1 26 132
95 3 1 25 115
96 3 2 28 164
97 3 6 22 110
98 2 6 20 149
99 1 5 28 161
end


quietly forvalues i = 20/30 & j = 1/6 { 
    gen x =  status [w=weight] if angle== `i' & group= `j', detail  
 }

I have data at one time point. I want to apply the sample weights "weight" to the "status" variable to get weighted counts of status for each combination of "angle" and "group". There are 11 angles and 6 groups and thus I basically want a new data set that is 66 rows and has 3 columns--for each status. Thank you

[Nick Cox]

Thanks for the data example. This seems confused:

1. If you want R code for this, Statalist is not the place to ask.

2. The code following your data example is a series of wild guesses. There are several reasons why it won't do what I think you want, but I will not detail them.

This code may help:

contract angle group status [fw=weight]
replace _freq = 0 if _freq == . 
reshape wide _freq, i(group angle) j(status)
sort angle group
list angle group _freq* , sepby(angle)