HTSeedGLM (version 0.1.0)

FStat: F-test between two fitted models

Description

This function considers two fitted models as inputs. Considering the first model as full model, it performs testing equality of uniformity parameters representing the model under null hypothesis and provides the p-value and degrees of freedom of the test statistic.

Usage

FStat(model1, model2)

Value

• Degrees of freedom and p-value

Arguments

model1

First fitted model

model2

Second fitted model

References

• Bradford, K. J. (2002). Applications of Hydrothermal Time to Quantifying and Modeling Seed Germination and Dormancy. Weed Science, 50(2), 248–260. http://www.jstor.org/stable/4046371

• Kebreab, E., & Murdoch, A. J. (1999). Modelling the effects of water stress and temperature on germination rate of Orobanche aegyptiaca seeds. Journal of Experimental Botany, 50(334), 655-664. doi:10.1093/jxb/50.334.655

• Dobson, A. J., & Barnett, A. G. (2018). An introduction to generalized linear models. Chapman and Hall/CRC.

Examples

data1 <- data.frame(cbind(sg = c(rep(1, 95), rep(0, 5), rep(1, 87), rep(0, 13),
rep(1, 80), rep(0, 20), rep(1, 59), rep(0, 41),
rep(1, 50), rep(0, 50), rep(1, 79), rep(0, 21),
rep(1, 69), rep(0, 31), rep(1, 72), rep(0, 28),
rep(1, 44), rep(0, 56), rep(1, 14), rep(0, 86)),
v1 = c(rep(1, 500), rep(0, 500)),
v2 = c(rep(0, 500), rep(1, 500)),
wp1 = c(rep(0, 100), rep(-0.3, 100), rep(-0.6, 100),
       rep(-0.9, 100), rep(-1.2, 100), rep(0, 500)),
wp2 = c(rep(0, 600), rep(-0.3, 100), rep(-0.6, 100),
       rep(-0.9, 100), rep(-1.2, 100))))
data2 <- data.frame(cbind(sg = c(rep(1, 95), rep(0, 5), rep(1, 87), rep(0, 13),
rep(1, 80), rep(0, 20), rep(1, 59), rep(0, 41),
rep(1, 50), rep(0, 50), rep(1, 79), rep(0, 21),
rep(1, 69), rep(0, 31), rep(1, 72), rep(0, 28),
rep(1, 44), rep(0, 56), rep(1, 14), rep(0, 86)),
v1 = c(rep(1, 500), rep(0, 500)),
v2 = c(rep(0, 500), rep(1, 500)),
wp = c(rep(0, 100), rep(-0.3, 100), rep(-0.6, 100),
      rep(-0.9, 100), rep(-1.2, 100), rep(0, 100),
      rep(-0.3, 100), rep(-0.6, 100), rep(-0.9, 100),
      rep(-1.2, 100))))
myprobit1 <- glm(sg ~ v1 + v2 + wp1 + wp2 - 1, data = data1, family = binomial(link = probit))
myprobit2 <- glm(sg ~ v1 + v2 + wp - 1, data = data2,
family = binomial(link = probit))
my.f<- FStat(myprobit1, myprobit2)