EXAMPLE SAS PROGRAMS FOR ESTIMATING GENOTYPIC AND PHENOTYPIC CORRELATIONS AND THEIR STANDARD ERRORS

 

Example SAS codes correspond to J.B. Holland, "Estimating genotypic correlations and their standard errors using multivariate restricted maximum likelihood estimation in SAS Proc MIXED" 2005, Crop Science (Accepted). Preprints of the manuscript available on request: james_holland@ncsu.edu.

All examples use data from Holland et al. (2002), available in an excel spreadsheet.

Choose from four different examples for genotypic and phenotypic correlation estimation using REML:

1. Correlations estimated from randomized complete block designs in multiple environments. (Equivalent to code presented in manuscript).

2. Correlations estimated from incomplete block designs in multiple environments.

3. Correlations estimated from a randomized complete block design in one environment.

4. Correlations estimated from an incomplete block design in one environment.

 

SAS programs for simulation study described in manuscript, to create simulation data sets, to randomly sample missing data, and to analyze data sets with MANOVA and REML.

back to jim's homepage