Description Usage Format Details Source Examples

Yields of wheat cultivars introduced 1860-1982. Grown in 20 environments.

1 | ```
data("perry.springwheat")
``` |

A data frame with 560 observations on the following 6 variables.

`yield`

yield, kg/ha

`gen`

genotype/cultivar factor, 28 levels

`env`

environment factor, 20 levels

`site`

site factor

`year`

year, 1979-1982

`yor`

year of release, 1860-1982

Twenty-eight of the most significant wheat cultivars of the past century in Western Australia, were grown in 20 field trials over 4 years in the Central and Eastern wheat-belt of Australia.

At the Wongan Hills site there were separate early and late sown trials in 1979 and 1980. Later sowing dates generally have lower yields.

Note: Although not indicated by the original paper, it may be that the Merredin site in 1979 also had early/late sowing dates.

MW Perry and MF D'Antuono. (1989).
Yield improvement and associated characteristics of some Australian
spring wheat cultivars introduced between 1860 and 1982.
*Australian Journal of Agricultural Research*, 40(3), 457–472.
http://www.publish.csiro.au/nid/43/issue/1237.htm

Used with permission of Mario D'Antuono and CSIRO Publishing.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 | ```
data(perry.springwheat)
dat <- perry.springwheat
library(lattice)
xyplot(yield~yor|env, dat, type=c('p','r'), xlab="year of release",
main="perry.springwheat")
# Show a table of sites*year
# library(latticeExtra)
# useOuterStrips(xyplot(yield~yor|site*factor(year), dat,
# type=c('p','r')))
# Perry reports a rate of gain of 5.8 kg/ha/year. No model is given.
# We fit a model with separate intercept/slope for each env
m1 <- lm(yield ~ env + yor + env:yor, data=dat)
# Average slope across environments
mean(c(coef(m1)[21], coef(m1)[21]+coef(m1)[22:40]))
## [1] 5.496781
# ----------------------------------------------------------------------------
## Not run:
# Now a mixed-effects model. Fixed overall int/slope. Random env int/slope.
# First, re-scale response so we don't have huge variances
dat$y <- dat$yield / 100
require(lme4)
# Use || for uncorrelated int/slope. Bad model. See below.
# m2 <- lmer(y ~ 1 + yor + (1+yor||env), data=dat)
## Warning messages:
## 1: In checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.55842 (tol = 0.002, component 1)
## 2: In checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model is nearly unidentifiable: very large eigenvalue
## - Rescale variables?;Model is nearly unidentifiable: large eigenvalue ratio
## - Rescale variables?
# Looks like lme4 is having trouble with variance of intercepts
# There is nothing special about 1800 years, so change the
# intercept -- 'correct' yor by subtracting 1800 and try again.
dat$yorc <- dat$yor - 1800
m3 <- lmer(y ~ 1 + yorc + (1+yorc||env), data=dat)
# Now lme4 succeeds. Rate of gain is 100*0.0549 = 5.49
fixef(m3)
## (Intercept) yorc
## 5.87492444 0.05494464
# asreml3
require(asreml)
m3a <- asreml(y ~ 1 + yorc, data=dat, random = ~ env + env:yorc)
require(lucid)
vc(m3)
## grp var1 var2 vcov sdcor
## env (Intercept) <NA> 11.61 3.407
## env.1 yorc <NA> 0.00063 0.02511
## Residual <NA> <NA> 3.551 1.884
vc(m3a)
## effect component std.error z.ratio con
## env!env.var 11.61 4.385 2.6 Positive
## env:yorc!env.var 0.00063 0.000236 2.7 Positive
## R!variance 3.551 0.2231 16 Positive
## End(Not run)
# ----------------------------------------------------------------------------
``` |

agridat documentation built on May 2, 2019, 4:01 p.m.

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.