第二章简单线性回归总论.pptVIP

回归系数的Monte Carlo模拟(R) x-c(9.7,10.1,10.0,10.4,10.1,10.2,9.7,10.4,9.6,9.8) b1-numeric(2000) set.seed(20) for(i in 1:2000){ y-2+5*x+rnorm(10,0,5) b1[i]-coef(lm(y~x))[[2]]} hist(b1) mean(b1);sd(b1) sqrt(25/sum((x-mean(x))^2))# require(tseries) jarque.bera.test(b1)#正态性检验 回归系数的Monte Carlo模拟(Gretl) nulldata 10 set seed 2012 loop 2000 --progressive --quiet series x={9.7,10.1,10.0,10.4,10.1,10.2,9.7,10.4,9.6,9.8} genr y1 = 2 + 5*x + 5*normal() ols y1 const x genr b2 = $coeff(x) print b2 store d:\coeff.gdt b2 endloop open d:\coeff.gdt normtest b2 --jbera 教材案例分析程序 create P38 a 1980 1998 read D:\Econom