agridat (version 1.27)

petersen.wheat: Augmented design of durum wheat

Description

Augmented design of durum wheat with 30 new selections and 3 check varieties.

Usage

data("petersen.wheat")

Format

A data frame with 48 observations on the following 4 variables.

plot

Plot number, 1-48

block

Block factor, 6 levels

gen

Genotype factor, 33 levels

yield

Yield, kg/ha

Details

A cereal breeder conducted a preliminary yield trial on 30 new selections of durum wheat. The new selections were compared against three check (standard) varieties: St (Stork), Ci (Cimmaron), and Wa (Waha).

Only enough seed of the new varieties was available to plant a single row of each, so an augmented design was used with 6 blocks. Each block contains 8 plots: 3 check varieties (replicated in every block) plus 5 unreplicated new selections.

The field layout consisted of a single line of 48 plots.

Petersen reports MSE = 91,103 from the ANOVA of check varieties.

Data collected for agridat by Paul Schmidt.

References

None

Examples

if (FALSE) {

library(agridat)
data(petersen.wheat)
dat <- petersen.wheat

libs(desplot)

# True layout: single line of 48 plots
# For visualization, rearrange into a grid (8 rows x 6 columns)
dat <- transform(dat,
  row = (plot - 1) %% 8 + 1,
  col = (plot - 1) %/% 8 + 1
)

# Custom colors: green shades for 30 new selections, orange shades for 3 checks
checks <- c("Ci", "St", "Wa")
new_gens <- sort(setdiff(levels(dat$gen), checks))
col_checks <- colorRampPalette(c("darkorange", "orange", "goldenrod"))(3)
col_new <- colorRampPalette(c("darkgreen", "forestgreen", "limegreen"))(30)
names(col_checks) <- checks
names(col_new) <- new_gens
gen_cols <- c(col_checks, col_new)

desplot(dat, gen ~ col*row,
  col.regions = gen_cols,
  out1 = block,
  text = gen, shorten = "no", cex = 0.7,
  show.key = FALSE,
  flip = TRUE,
  main = "petersen.wheat - rearranged as grid"
)

}