在R中使用glm(..)获得95％的置信区间 [英] Get 95% confidence interval with glm(..) in R

问题描述

这是一些数据

dat = data.frame(y = c(9,7,7,7,5,6,4,6,3,5,1,5), x = c(1,1,2,2,3,3,4,4,5,5,6,6), color = rep(c('a','b'),6))

以及这些数据的绘图

require(ggplot)
ggplot(dat, aes(x=x,y=y, color=color)) + geom_point() + geom_smooth(method='lm')

使用功能 MCMCglmm()…

require(MCMCglmm)
summary(MCMCglmm(fixed = y~x/color, data=dat))

我得到了估计值的上下95％的间隔，使我知道两个斜率(颜色= a和颜色= b)是否显着不同.

I get the lower and upper 95% interval for the estimate allowing me to know if the two slopes (color = a and color = b) are significantly different.

查看此输出时...

summary(glm(y~x/color, data=dat))

...我看不到置信区间！

... I can't see the confidence interval!

我的问题是:

使用函数 glm()时，如何对估计值具有上下95％的置信度?

How can I have these lower and upper 95% interval confidence for the estimates when using the function glm()?

推荐答案

使用 confint

mod = glm(y~x/color, data=dat)
summary(mod)
Call:
glm(formula = y ~ x/color, data = dat)

Deviance Residuals: 
     Min        1Q    Median        3Q       Max  
-1.11722  -0.40952  -0.04908   0.32674   1.35531  

Coefficients:
            Estimate Std. Error t value     Pr(>|t|)
(Intercept)   8.8667     0.4782  18.540 0.0000000177
x            -1.2220     0.1341  -9.113 0.0000077075
x:colorb      0.4725     0.1077   4.387      0.00175

(Dispersion parameter for gaussian family taken to be 0.5277981)

    Null deviance: 48.9167  on 11  degrees of freedom
Residual deviance:  4.7502  on  9  degrees of freedom
AIC: 30.934

Number of Fisher Scoring iterations: 2

confint(mod)
Waiting for profiling to be done...
                 2.5 %     97.5 %
(Intercept)  7.9293355  9.8039978
x           -1.4847882 -0.9591679
x:colorb     0.2614333  0.6836217

这篇关于在R中使用glm(..)获得95％的置信区间的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！