Uniformity trial of cotton in South Africa

Usage

data("macdonald.cotton.uniformity")

Format

A data frame with 1152 observations on the following 4 variables.

row: row ordinate
col: column ordinate
yield: yield, grams per plot
field: field identifier

Details

Experiments conducted at Cotton Experiment Station, Barberton, South Africa 1935-1936. Planted Dec 6. Yield was about 400 pounds per acre. (Plot yields are given in grams). Yields were low due to late planting 6-8 weeks. Severe damage by aphids unevenly. Severe drought in Feb/Mar. Both fields represent extreme variation in soil fertility at the station and are not typical.

Field A: 144 rows 3 ft 6 in apart, 18 in between plants. Rows divided into 30 ft units. In Rothamsted documents, this field is called "B5b".

Field B: 144 rows 3 ft 6 in apart, 2 ft between plants. Rows divided into 40 ft units. In the Rothamsted documents, this field is called "B2a".

This data was made available with special help from the staff at Rothamsted Research Library. Data were typed by K.Wright.

Source

Rothamsted Research Library, Box STATS17 WG Cochran, Folder 8.

References

MacDonald, D. and Fielding, W. L. and Ruston, D. F. (1939). Experimental methods with cotton: I. The design of plots for variety trials. The Journal of Agricultural Science, 29, 35-47. http://dx.doi.org/10.1017/S0021859600051534 https://archive.org/details/in.ernet.dli.2015.26648/page/n45/

Examples

if (FALSE) { # \dontrun{
  library(agridat)
  data(macdonald.cotton.uniformity)
  dat <- macdonald.cotton.uniformity

  libs(desplot)
  desplot(dat, yield ~ col*row, subset=field=="A",
          flip=TRUE, tick=TRUE, aspect=(144*3.5)/(4*30),
          main="macdonald.cotton.uniformity - field A")
  desplot(dat, yield ~ col*row, subset=field=="B",
          flip=TRUE, tick=TRUE, aspect=(144*3.5)/(4*40),
          main="macdonald.cotton.uniformity - field B")

  # Match MacDonald 1939 Figure 1
  #libs(dplyr)
  #d2 <- subset(dat, field=="B")
  #d2 <- mutate(d2, block=rep(rep(1:36, each=4), 4))
  #d2 <- group_by(d2, block, col)
  #d2 <- summarize(d2, yield=mean(yield))
  #d2 <- ungroup(d2)
  #desplot(d2, yield~col*block, flip=TRUE)
} # }