Stata Technical Bulletin
11
I marital
|
inpovI |
0 |
1 |
2 |
3 |
4 I |
Total |
|
0 I |
12630 94.42 |
87ε 78.83 |
1926 81.22 |
322 75.06 |
8960 I 81.28 I |
— 24712 87.30 |
|
i I |
746 |
23ε 21.17 |
446 18.78 |
107 24.94 |
2063 I 18.72 I |
— 3596 12.70 |
|
-----------+— Total I |
13376 100.00 |
1110 100.00 |
2370 100.00 |
429 100.00 |
11023 I 100.00 I |
28308 100.00 |
. tabulate race [weight=pwgt]
(frequency weights assumed)
race I Freq. Percent Cum.
|
— |
-+----------------------------------- | ||
|
1 |
I 26059 |
92.06 |
92.06 |
|
2 |
I 530 |
1.87 |
93.93 |
|
4 |
I 19 |
0.07 |
93.99 |
|
6 |
I 209 |
0.74 |
94.73 |
|
8 |
I 77 |
0.27 |
95.00 |
|
9 |
I 93 |
0.33 |
95.33 |
|
10 |
I 28 |
0.10 |
95.43 |
|
11 |
I 226 |
0.80 |
96.23 |
|
12 |
I 54 |
0.19 |
96.42 |
|
14 |
I 19 |
0.07 |
96.49 |
|
15 |
I 16 |
0.06 |
96.55 |
|
16 |
I 51 |
0.18 |
96.73 |
|
23 |
I 38 |
0.13 |
96.86 |
|
37 |
I 516 |
1.82 |
98.68 |
|
302 |
I 10 |
0.04 |
98.72 |
|
304 |
I 13 |
0.05 |
98.76 |
|
307 |
I 6 |
0.02 |
98.78 |
|
308 |
I 6 |
0.02 |
98.81 |
|
312 |
I 16 |
0.06 |
98.86 |
|
315 |
I 3 |
0.01 |
98.87 |
|
316 |
I 29 |
0.10 |
98.98 |
|
317 |
I 22 |
0.08 |
99.05 |
|
320 |
I 7 |
0.02 |
99.08 |
|
322 |
I 12 |
0.04 |
99.12 |
|
323 |
I 57 |
0.20 |
99.32 |
|
326 |
I 131 |
0.46 |
99.78 |
|
327 — |
I 61 .+--------------- |
0.22 |
100.00 |
Total I 28308 100.00
. * where :
. * 001 White 002 Black 004 Eskimo 005 Aleut
. * ...
. * others are as defined in the PUMS documentation.
As you can see, the PUMS data provide a rich source of socioeconomic data that can readily be analyzed using Stata. We
encourage you to explore these data.
sg26.1 Fractional polynomials: Correction
Patrick Royston, Royal Postgraduate Medical School, London, FAX (011)-44-81-740-3119
Douglas G. Altman, Imperial Cancer Research Fund, London, FAX (011)-44-71-269-3429
Three errors in fp.ado have come to light since its publication in STB-21 (Royston and Altman 1994). These are corrected
in the present release. The errors are as follows.
1. The comparison and estimates options were inoperative when used at the model-fitting stage (they did work in ‘replay’
mode).
2. Replay of models containing basevars did not display regression coefficients for the basevars as it should.
3. With the estimates option, the deviance of the best-fitting fractional polynomial (FP) model should be displayed at the
foot of the regression output. In fact, the deviance of the base model (that is, without FP terms) was given instead.