toMappedDataGroupedByModelInputNames

Cached simulation runs are associated with instances of ModelInputs. The equality of model inputs is determined by the model identifier, the names (and values) of the inputs, and the names of the desired responses. For the purposes of this function, we assume that the simulation runs are grouped (partitioned) by those model inputs that have the same input names and same requested responses. See simulationRunsGroupedByModelInputNames(). This function returns the inputs and replication responses within a map of maps. The outer map represents the common input names and requested responses, and the inner map the inputs and responses. You can think of the inner map as holding the columns of the simulation run inputs and responses for each replication within the run. This "tabular" representation is useful for analysis of input to output relationships. The model input values will be repeated for each replication. The response values represent the observed replication average for the response. The outer grouping is necessary because not every requested simulation run will have the same input names and/or requested responses. The grouping ensures those that have common names (inputs/response) will be within the associated map.