agridat (version 1.5)

ortiz.tomato: Tomato weight/yield and environmental covariates in Latin America

Description

Tomato weight/yield and environmental covariates in Latin America

Usage

data(ortiz.tomato)

format

A list of three matrices, yield, weight, and covs. See details below.

source

Rodomiro Ortiz and Jose Crossa and Mateo Vargas and Juan Izquierdo, 2007. Studying the Effect of Environmental Variables On the Genotype x Environment Interaction of Tomato. Euphytica, 153, 119--134. Used with permission of Rodomiro Ortiz.

Details

The yield matrix contains average marketable fruit yield (t / ha) for 8 open-pollinated and 7 hybrid tomatos in 18 environments. The weight matrix contains average fruit weight (g). The environment locations and codes are: Estanzuela, Guatemala (E04), Baja Verapaz, Guatemala, (E05), Cogutepeque, El Salvador (E06), San Andres, El Salvador (E07), Comayagua, Honduras (E11), Valle de Sabaco, Nicaragua (E14), San Antonio de Belen, Costa Rica (E15), San Cristobal, Dominican Republic (E20), Constanza, Dominican Republic (E21), Palmira, Colombia (E27), La Molina, Peru (E40), Santiago, Chile (E41), Chillan, Chile (E42), Curacavi, Chile (E43), Colina, Chile (44), Belem, Brazil (E50), Caacupe, Paraguay (E51), Centeno, Trinidad Tobago (E53). The covs matrix contains 16 environmental covariates. ll{ Lat Latitude Long Longitude MxT Max temperature (C) MnT Min temperature (C) MeT Mean temperature (C) Prec Precipitation (mm) Day Degree days (base 10) pH Soil pH OM Organic matter (percent) P Phosphorous (ppm) K Potassium (me/100 g) ExN Extra nitrogen (kg / ha) ExP Extra phosphorous (kg / ha) ExK Extra potassium (kg / ha) Trim Trimming (0/1) Driv Drivings (0/1) Irr Irrigation (0/1) Dha Days to harvest }

Examples

# Double-centered yield matrix
Y <- ortiz.tomato$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
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
require(pls)
m1 <- plsr(Y~X)
# Inner-product relationships similar to Ortiz figure 1.
biplot(m1, which="x", var.axes=TRUE)
biplot(m1, which="y", var.axes=TRUE)