Compute a list of covariance matrices corresponding to the "Unassociated", "Directly associated" and "Indirectly associated" models

cov_udi(data, model = udi_model_matrix(n_conditions(data)))

Arguments

data: a mash data object, eg as created by mash_set_data
model: a model matrix with R columns, where R is the number of conditions in the data; each row should be a vector of length R with elements "U","D" and "I" indicating whether each effect is Unassociated, Directly associated or Indirectly associated

Value

a named list of covariance matrices

Details

If model is specified then this returns the covariance matrices for those models. The default creates all possible models. For a desription of the "Unassociated", "Directly associated" and "Indirectly associated" models see Stephens M (2013), A unified framework for Association Analysis with Multiple Related Phenotypes, PloS ONE.

Examples

data = mash_set_data(Bhat = cbind(c(1,2),c(3,4)), Shat = cbind(c(1,1),c(1,1)))
cov_udi(data)
#> $cov_udi_DU
#>      [,1] [,2]
#> [1,]    1    0
#> [2,]    0    0
#> 
#> $cov_udi_UD
#>      [,1] [,2]
#> [1,]    0    0
#> [2,]    0    1
#> 
#> $cov_udi_DD
#>      [,1] [,2]
#> [1,]    1    0
#> [2,]    0    1
#> 
#> $cov_udi_ID
#>      [,1] [,2]
#> [1,]    0    0
#> [2,]    0    1
#> 
#> $cov_udi_DI
#>      [,1] [,2]
#> [1,]    1    0
#> [2,]    0    0
#> 
cov_udi(data,c('I','D'))
#> $cov_udi_ID
#>      [,1] [,2]
#> [1,]    0    0
#> [2,]    0    1
#>