Empty outputs now return data.tables with the corresponding column names with 0 rows instead of defaulting to lists of length 0
Column names for the outputs of many functions have changed in this release to be more standardized. Please refer to the function documentation. As a general rule names use camelCase separated by .s to indicate properties of a specific entity (eg. experiment.sampleCount) with the exception of acronyms which are always capitalized (experiment.ID)
update_results function added which allows re-creation of outputs of gemma.R functions without relying on the original code.
get_result_sets function added which allows accessing result sets directly, filtering them based on certain filterable properties (see filter_properties()$resultSet).
gemma_memoise function added which allows setting memoisation options without manually setting options
get_child_terms function added which returns child terms of an ontology term as inferred by Gemma
gemma_kable is added which returns tables formatted to fit
gemma.R 2.0.0
Breaking change to get_dataset_differential_expression_analyses function in order to return annotations for contrasts with multiple characteristics.