12 1 15 1 ;

PROC GENMOD;

MODEL y=x/DIST=POISSON LINK=id INTERCEPT=7.5 ITPRINT; Run;

T 7_3解答 DATA ex7_3;

INPUT y quality yanzheng age center f; IF quality = 1 ? yanzhneg=1 THEN bingsi=1; IF quality = 2 ? yanzhneg=1 THEN bingsi=2; IF quality = 1 ? yanzhneg=2 THEN bingsi=3; IF quality = 2 ? yanzhneg=2 THEN bingsi=4; CARDS;

1 1 1 1 1 9 1 1 2 1 1 4 ……

0 2 1 3 3 11 0 2 2 3 3 1 ;

PROC GENMOD; FREQ f;

INITIAL=5.0 MODEL y=bingsi age center/LINK=LOG DIST=NORMAL; PROC GENMOD; FREQ f;

MODEL y=bingsi age center/LINK=LOG DIST=BINORMIAL; PROC GENMOD; FREQ f;

MODEL y=bingsi age center/LINK=LOG DIST=POISSON; RUN;

T7-4解答 DATA ex7_4; INPUT r n x1 x2; a1=log(x1+1); a2=log(x2+1); t=a1*a2; CARDS; 11 200 0 0 ……

193 200 100 100 ;

PROC LOGISTIC; MODEL r/n=a1 a2 t; RUN; DATA a; SET ex7_4; Ln=log(n); PROC GENMOD;

MODEL r=a1 a2 t/DIST=POISSON OFFSET=ln; RUN;

负二项回归SAS不能解决,建议使用stata.

例7．5

TITLE’Poisson 回归’;

OPTION LS=150; DATA EX7_5;

INPUT d age sex month year @@; _pi=301415926;

M1=cos(_pi/6*month); M2=sin(_pi/6*month); Datalines;

0 0 1 1 78 1 0 1 1 81 ……..

4 2 0 12 80 2 2 0 12 83 ;

Proc genmod;

Model d=age sex/DIST=POISSON LINK =LOG; RUN;

PROC GENMOD;.

MODEL d=age sex M1 M2/DIST=POISSON LINK=LOG; RUN;

负二项回归SAS 不能解决，建议使用stata。

练习8.1

?0.65ts(t)?e（1）生存函数：曲线如下：

s(t)?e?0.65t

t??1

（3）生存大于两个单位时间的概率：

??1.5385tm?ln0.5?1.0664?0.65

s(t)t?2?e?0.65tt?2?e?0.65?2?0.27253719?(?t)m

8.2（1）生存函数：s?t??e?e?(0.8t)3曲线如下：

m?1 危险度函数：h(t)?m?(?t)?3?0.8(0.8t)2