Purpose
To work on exercises from Dobson’s Book.

  1. Hiroshima Exercise
> folder <- "C:/Cauldron/garage/R/soulcraft/Volatility/Learn/Dobson-GLM/"
> file.input <- paste(folder, "Table 7.11 Hiroshima deaths.csv",
+     sep = "")
> data <- read.csv(file.input, header = T, stringsAsFactors = F)
> data <- data[1:6, ]
> data$prob <- data[, 2]/(data[, 2] + data[, 3])
> data$radiation <- factor(data$radiation)
> plot(1:6, data$prob, pch = 19, col = "blue")

Chap-7-Exer-1-002.jpg

Plot Log odds

> plot(1:6, log(data$prob/(1 - data$prob)), pch = 19, col = "blue")
  1. Ok I took a peak at the solution and author uses lower bound of interval
> data$dose <- c(0, 1, 10, 50, 100, 200)
  1. Use Logit
> fit <- glm(cbind(data$leukemia, data$other.cancer) ~ data$dose,
+     family = binomial(link = "logit"))
> summary(fit)
Call:
glm(formula = cbind(data$leukemia, data$other.cancer) ~ data$dose,
    family = binomial(link = "logit"))
Deviance Residuals: 1 2 3 4 5 6 0.4142808 -0.4899417 -0.1399059 0.0283528 0.0004824 0.0026940
Coefficients: Estimate Std. Error z value Pr(>|z|) (Intercept) -3.488973 0.204062 -17.098 < 2e-16 *** data$dose 0.014410 0.001817 7.932 2.15e-15 *** --- Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
(Dispersion parameter for binomial family taken to be 1)
Null deviance: 54.35089 on 5 degrees of freedom Residual deviance: 0.43206 on 4 degrees of freedom AIC: 26.097
Number of Fisher Scoring iterations: 4