cbb {hnp} | R Documentation |
Coffee berry borer trapping data
Description
Data on counts of coffee berry borer obtained using different traps through time.
Usage
data(cbb)
Format
A data frame with 288 observations on the following 4 variables.
week | numeric | week of observation (1 to 24) |
block | factor | levels I II III IV |
trap | factor | levels CV F SF |
count | numeric | number of observed insects |
Details
The coffee berry borer is a major pest of commercial coffee. The insect directly attacks the coffee fruit in development causing severe losses in bean production and quality.
This data set was obtained in an experiment conducted by Mota (2013), where three types of traps (SF
, F
, CV
) were randomized in each of four equidistant lines (blocks) of a coffee field. Each week, over a 24 week period, the insects were removed from the traps and counted.
Source
Demétrio, C. G. B., Hinde, J. and Moral, R. A. (2014) Models for overdispersed data in entomology. In Godoy, W. A. C. and Ferreira, C. P. (Eds.) Ecological modelling applied to entomology. Springer.
References
Mota, L. H. C. (2013) Desenvolvimento de armadilha de auto-inoculacao para o controlde de Hypothenemus hampei (Ferrari, 1867) (Coleoptera: Curculionidae) com Beauveria bassiana (Bals.) Vuil (Ascomycota: Hypocreales) em tecido sinetico. Master's dissertation, ESALQ-USP
Examples
data(cbb)
# exploratory plot
require(latticeExtra)
trellis.par.set(strip.background=list(col="lightgrey"))
useOuterStrips(xyplot(count ~ week | block + trap, data=cbb,
layout=c(3,1),type="l", col=1, xlab="Week", ylab="Insect counts"))
# Poisson fit
model1 <- glm(count ~ block + trap*factor(week),
data=cbb, family=poisson)
anova(model1, test="Chisq")
sum(resid(model1, ty="pearson")^2)
summary(model1)
hnp(model1, sim=19, conf=1)
## Not run:
hnp(model1) # default call
## End(Not run)
# Quasi-Poisson fit
model2 <- glm(count ~ block + trap*factor(week), data=cbb,
family=quasipoisson)
anova(model2, test="F")
summary(model2)
hnp(model2, sim=19, conf=1)
## Not run:
hnp(model2) # default call
## End(Not run)
## for discussion on the analysis of this data set,
## see Demetrio et al. (2014)