Learn R Programming

agridat (version 1.23)

ortiz.tomato: Multi-environment trial of tomato in Latin America, weight/yield and environmental covariates

Description

Multi-environment trial of tomato in Latin America, weight/yield and environmental covariates

Usage

data("ortiz.tomato.covs")
  data("ortiz.tomato.yield")

Arguments

Format

The ortiz.tomato.covs data frame has 18 observations on the following 18 variables.

env

environment

Day

degree days (base 10)

Dha

days to harvest

Driv

drivings (0/1)

ExK

extra potassium (kg / ha)

ExN

extra nitrogen (kg / ha)

ExP

extra phosphorous (kg / ha)

Irr

irrigation (0/1)

K

potassium (me/100 g)

Lat

latitude

Long

longitude

MeT

mean temperature (C)

MnT

min temperature (C)

MxT

max temperature (C)

OM

organic matter (percent)

P

phosphorous (ppm)

pH

soil pH

Prec

precipitation (mm)

Tri

trimming (0/1)

The ortiz.tomato.yield data frame has 270 observations on the following 4 variables.

env

environment

gen

genotype

yield

marketable fruit yield t/ha

weight

fruit weight, g

Details

The environment locations are:

E04Estanzuela, Guatemala
E05Baja Verapaz, Guatemala
E06Cogutepeque, El Salvador
E07San Andres, El Salvador
E11Comayagua, Honduras
E14Valle de Sabaco, Nicaragua
E15San Antonio de Belen, Costa Rica
E20San Cristobal, Dominican Republic
E21Constanza, Dominican Republic
E27Palmira, Colombia
E40La Molina, Peru
E41Santiago, Chile
E42Chillan, Chile
E43Curacavi, Chile
E44Colina, Chile
E50Belem, Brazil
E51Caacupe, Paraguay
E53Centeno, Trinidad Tobago

Used with permission of Rodomiro Ortiz.

Examples

Run this code
if (FALSE) {

library(agridat)
data(ortiz.tomato.covs)
data(ortiz.tomato.yield)

libs(pls, reshape2)
# Double-centered yield matrix
Y <- acast(ortiz.tomato.yield, env ~ gen, value.var='yield')
Y <- sweep(Y, 1, rowMeans(Y, na.rm=TRUE))
Y <- sweep(Y, 2, colMeans(Y, na.rm=TRUE))

# Standardized covariates
X <- ortiz.tomato.covs
rownames(X) <- X$env
X <- X[,c("MxT", "MnT", "MeT", "Prec", "Day", "pH", "OM", "P", "K",
          "ExN", "ExP", "ExK", "Trim", "Driv", "Irr", "Dha")]
X <- scale(X)

# Now, PLS relating the two matrices.
# Note: plsr deletes observations with missing values

m1 <- plsr(Y~X)
# Inner-product relationships similar to Ortiz figure 1.
biplot(m1, which="x", var.axes=TRUE, main="ortiz.tomato - env*cov biplot")
#biplot(m1, which="y", var.axes=TRUE)
}

Run the code above in your browser using DataLab