Open PavelBal opened 7 months ago
sensitivityCalculation( simulation, outputPaths, parameterPaths, variationRange = list(range1,range2), (same size than parameterPaths) variationType = list("absolute","relative"), (same size than parameterPaths) pkParameters = list("C_max", "t_max", "AUC_inf", "my function" = function), (warning: must be a named list) saOutputFilePath = NULL, simulationRunOptions = NULL )
Is variationRange
a list of numeric vectors, e.g., list(c(0.5, 1, 2), c(1,2,3))
? Does this mean that different variation parameters might have different variation ranges? Yes, doable, we can support this. We should pay attention to how this will be represented in the plots.
Then it should be possible to provide one vector that will be applied to all parameters. So variationRange = c(1,2,3)
means that this varation range (or absolute values) will be applied for all parameters. Otherwise, variationRange
must be of the same length as parameterPaths
.
Same applies to variationType
.
Use ospsuite.utils::toList
to ensure that variationRange
and variationType
are lists (and check for their length).
I propose to have the custom functions not as part of the pkParameters
argument (because the function is not a PK-parameter), but as a separate argument customOutputFunctions = NULL
. Must be a named list (check for names), each group entry must be a function with arguments SimulationResults, outputPaths
.
Define inputs and outputs.