agridat (version 1.5)

johnson.blight: Potato blight due to weather in Prosser, Washington

Description

Potato blight due to weather in Prosser, Washington

Usage

data(johnson.blight)

source

Johnson, D.A. and Alldredge, J.R. and Vakoch, D.L. (1996). Potato late blight forecasting models for the semiarid environment of south-central Washington. Phytopathology, 86, 480--484.

Details

The variable 'blight detected' is 1 if 'area' > 0.

References

Vinayanand Kandala, Logistic Regression.

Examples

data(johnson.blight)

dat <- johnson.blight

# Define indicator for blight in previous year
dat$blight.prev[2:25] <- dat$blight[1:24]
dat$blight.prev[1] <- 0 # Need this to match the results of Johnson
dat$blight.prev <- factor(dat$blight.prev)
dat$blight <- factor(dat$blight)

# Johnson et al developed two logistic models to predict outbreak of blight
m1 <- glm(blight ~ blight.prev + rain.am + rain.ja, data=dat, family=binomial)
m1
m2 <- glm(blight ~ blight.prev + rain.am + precip.m, data=dat, family=binomial)
m2