Missing data policy for ssGSEA, as per feature request in #168, it resolves #174. Checking for the presence of missing (NA) values is done through the function anyNA(). By default, such a check will only occur when the input expression data container is a base matrix, an ExpressionSet, or a SummarizedExperiment, otherwise it will be assumed that there are no missing values. This can be overridden with the argument checkNA="yes" or checkNA="no", at parameter construction. Once the missing values are identified, this PR currently propagates them through calculations (use="everything"), removes them (use="na.rm") or produces an error (use="all.obs"). If normalize=TRUE and missing values are overlooked, NA values may propagate to the normalising constant and this will produce an error.
Missing data policy for ssGSEA, as per feature request in #168, it resolves #174. Checking for the presence of missing (
NA) values is done through the functionanyNA(). By default, such a check will only occur when the input expression data container is a basematrix, anExpressionSet, or aSummarizedExperiment, otherwise it will be assumed that there are no missing values. This can be overridden with the argumentcheckNA="yes"orcheckNA="no", at parameter construction. Once the missing values are identified, this PR currently propagates them through calculations (use="everything"), removes them (use="na.rm") or produces an error (use="all.obs"). Ifnormalize=TRUEand missing values are overlooked,NAvalues may propagate to the normalising constant and this will produce an error.