This PR fixes the issue reported in #42, which caught an out-of-date use of t() on the object produced by SummarizedExperiment::assay(), which (at least as of Bioconductor 3.7) returns an object of class DataFrame. In biomarkertmle(), t() and other downstream functions (e.g., limma::normalizeBetweenArrays()) require a Matrix-like object, a status for which DataFrame apparently does not qualify. This has been fixed by adding an explicit call to as.matrix. The opportunity is also taken to amend documentation and make other less-intrusive packaging changes.
Changes
explicitly coerce object returned by assay() to class Matrix
documentation updates + updates to pkgdown site
names of .R files have been amended for the sake of clarity
explicitly require all of S4Vectors via @import to satisfy pkgdown
This PR fixes the issue reported in #42, which caught an out-of-date use of
t()
on the object produced bySummarizedExperiment::assay()
, which (at least as of Bioconductor 3.7) returns an object of classDataFrame
. Inbiomarkertmle()
,t()
and other downstream functions (e.g.,limma::normalizeBetweenArrays()
) require aMatrix
-like object, a status for whichDataFrame
apparently does not qualify. This has been fixed by adding an explicit call toas.matrix
. The opportunity is also taken to amend documentation and make other less-intrusive packaging changes.Changes
assay()
to classMatrix
pkgdown
site.R
files have been amended for the sake of clarityS4Vectors
via@import
to satisfypkgdown