データ解析のための統計モデリング入門6章

> summary(d)
       N           y              x          f     
 Min.   :8   Min.   :0.00   Min.   : 7.660   C:50  
 1st Qu.:8   1st Qu.:3.00   1st Qu.: 9.338   T:50  
 Median :8   Median :6.00   Median : 9.965         
 Mean   :8   Mean   :5.08   Mean   : 9.967         
 3rd Qu.:8   3rd Qu.:8.00   3rd Qu.:10.770         
 Max.   :8   Max.   :8.00   Max.   :12.440         
> logistic <- function(z) 1 / (1 + exp(-z)) #関数の定義
> z <- seq(-6, 6, 0.1)
> plot(z, logistic(z), type = "l")
> glm(cbind(y, N - y) ~ x + f, data = d, family = binomial)

Call:  glm(formula = cbind(y, N - y) ~ x + f, family = binomial, data = d)

Coefficients:
(Intercept)            x           fT  
    -19.536        1.952        2.022  

Degrees of Freedom: 99 Total (i.e. Null);  97 Residual
Null Deviance:	    499.2 
Residual Deviance: 123 	AIC: 272.2
> library(MASS) #stepAICを定義するMASSpackageを読み込み
> stepAIC(fit.xf)
Error in terms(object) : object 'fit.xf' not found
> fit.xf <- glm(cbind(y, N - y) ~ x + f, data = d, family = binomial)
> stepAIC(fit.xf)
Start:  AIC=272.21
cbind(y, N - y) ~ x + f

       Df Deviance    AIC
<none>      123.03 272.21
- f     1   217.17 364.35
- x     1   490.58 637.76

Call:  glm(formula = cbind(y, N - y) ~ x + f, family = binomial, data = d)

Coefficients:
(Intercept)            x           fT  
    -19.536        1.952        2.022  

Degrees of Freedom: 99 Total (i.e. Null);  97 Residual
Null Deviance:	    499.2 
Residual Deviance: 123 	AIC: 272.2
> glm(cbind(y, N - y) ~ x * f, family = binomial, data = d)

Call:  glm(formula = cbind(y, N - y) ~ x * f, family = binomial, data = d)

Coefficients:
(Intercept)            x           fT         x:fT  
  -18.52332      1.85251     -0.06376      0.21634  

Degrees of Freedom: 99 Total (i.e. Null);  96 Residual
Null Deviance:	    499.2 
Residual Deviance: 122.4 	AIC: 273.6