SpATS (version 1.0-16)

wheatdata: Wheat yield in South Australia

Description

A randomized complete block experiment of wheat in South Australia (Gilmour et al., 1997).

Usage

data("wheatdata")

Format

A data frame with 330 observations on the following 7 variables

yield

yield, numeric

geno

wheat variety, a factor with 107 levels

rep

replicate factor, 3 levels

row

row position, numeric

col

column position, a numeric vector

rowcode

a factor with 2 levels. This factor is related to the way the trial and the plots were sown (see Details).

colcode

a factor with 3 levels. This factor is related to the way the plots were trimmed to an assumed equal length before harvest (see Details).

Details

A near complete block design experiment. In this experiment, a total of 107 varieties were sown in 3 replicates in a near complete block design (3 varieties were sown twice in each replicate). Each replicate consisted of 5 columns and 22 rows. Accordingly, the field consists on 15 columns and 22 rows (i.e. a total of 330 plots or observations). Trimming was done by spraying the wheat with herbicide. The sprayer travelled in a serpentine pattern up and down columns. More specifically, the plots were trimmed according to four possible combinations: up/down, down/down, down/up, and up/up. The colcode factor describe this 4-phase sequence. With respect to the way the trial was sown, the procedure was done in a serpentine manner with a planter that seeded three rows at a time (Left, Middle, Right). This fact led to a systematic effect on the yield due to sowing. The variable rowcode describes the position of each plot in the triplet (Left, Middle, Right). Level 1 was used for the middle plot, and 2 for those on the right or on the left of the triplet (see Gilmour et al. (1997) for more details).

Examples

# NOT RUN {
library(SpATS)
data(wheatdata)
summary(wheatdata)
# }