agridat (version 1.16)

wiedemann.safflower.uniformity: Uniformity trial of safflower

Description

Uniformity trial of safflower at Farmington, Utah, 1960.

Usage

data("wiedemann.safflower.uniformity")

Format

A data frame with 1782 observations on the following 3 variables.

row

row

col

column

yield

yield, grams

Details

This trial was planted at University Field Station, Farmington, Utah, in 1960, on a plot of land about one half acre in size. The soil was not too uniform...the northern third of the field was clay and the rest was gravelly. Rows were planted 22 inches apart, 62 rows total, each row running the length of the field. Before harvest, 4 rows were removed from each side, and 12 feet was removed from each end. Each row was harvested in five-foot lengths, threshed, and the seed weighed to the nearest gram.

The northern third of the field had yields twice as high as the remaining part of the field because the soil had better moisture retention. The remaining part of the field had yields that were more uniform.

Wiedemann determined the optimum plot size to be about 8 basic plots. The shape of the plot was not very important. But, two-row plots were recommended for simplicity of harvest, so 3.33 feet by 20 feet.

Based on operational costs, K1=74 percent and K2=26 percent.

Field width: 33 plots/ranges * 5ft = 165 feet

Field length: 54 rows * 22 in/row = 99 feet

For this R package, the tables in Wiedemann were converted by OCR to digital format, and all values were checked by hand.

The original source document has columns labeled 33, 32, ... 1. Here the columns are labeled 1:33 so that plotting tools work normally. See Wiedemann figure 8.

Wiedemann notes the statistical analysis of the data required 100 hours of labor. Today the analysis takes only a second.

References

None.

Examples

# NOT RUN {
  data(wiedemann.safflower.uniformity)
  dat <- wiedemann.safflower.uniformity

  # CV of entire field = 39<!-- %, matches Wiedemann p. 16 -->
  sd(dat$yield)/mean(dat$yield)

  require(desplot)
  desplot(yield~col*row, dat,
          flip=TRUE, tick=TRUE, aspect =99/165, # true aspect
          main="wiedemann.safflower.uniformity (true shape)")

  require(agricolae)
  require(reshape2)
  dmat <- acast(dat, row~col, value.var='yield')
  index.smith(dmat,
              main="wiedemann.safflower.uniformity",
              col="red")
  
# }