Learn R Programming

agridat (version 1.8.1)

wiebe.wheat.uniformity: Uniformity trial of wheat

Description

A uniformity trial of 1500 plots of wheat conducted in Idaho in 1927.

Arguments

source

Wiebe, G.A. 1935. Variation and Correlation in Grain Yield among 1,500 Wheat Nursery Plots, Journal of Agricultural Research, 50, 331-357. http://naldc.nal.usda.gov/download/IND43968632/PDF

Details

Yield trial conducted in 1927 near Aberdeen, Idaho. The crop was Federation wheat (C.I. no 4734). Plots were seeded on April 18 with a drill that sowed eight rows at a time. Individual rows were harvested in August and threshed with a small nursery thresher. Rows were 15 feet long, 1 foot apart. Some authors recommend analyzing the square root of the yields. Produced by the U.S. Dept of Agriculture.

References

D.A. Preece, 1981, Distributions of final digits in data, The Statistician, 30, 31--60.

Examples

Run this code
dat <- wiebe.wheat.uniformity

require("lattice")
# 125 rows = 125 feet tall. 12 cols * 15 feet = 180 feet wide.
desplot(yield~col+row, dat, aspect=125/180, flip=TRUE,
  main="Plot yields of wheat uniformity trial") # row 1 is at south

# Preece (1981) found the last digits have an interesting distribution.
dig <- substring(dat$yield, nchar(dat$yield))
dig <- as.numeric(dig)
hist(dig, breaks=0:10-.5, main="Histogram of last digit")
table(dat$col, dig) # Table 3 of Preece

# Median yields of rows/cols show obvious trends
plot(tapply(dat$yield, dat$row, median), xlab="Row", ylab="Yield")
plot(tapply(dat$yield, dat$col, median), xlab="Column", ylab="Yield")
# Spatial trends are obvious
xyplot(yield~col, dat, type=c('p','smooth'))

# Loess
m3 <- loess(yield~row+col, dat)
plot(fitted(m3), resid(m3), ylim=c(-300,300))

Run the code above in your browser using DataLab