MwPharm++/Edsim++ Release 1.8.2 - Description

• Added: Extended RxODE code generation templates (TMonitor, TIntegrator, TPatient, XCortisol). This is still experimental.
• Added: Added ID property to FitAlgorithm in order to uniquely identify multiple models in PopKin module.
• Added: Implemented parallel processing (fitting) of patients in PopKin module
• Added: PopKin module does not throw an abort exception anymore. Simply check the global Abort and Error flags on completion
• Added: Do not enforce maximum value of 1.0 on bioavailability F (all TInputEV descendants)
• Added: Added burst fraction to XInputController (Fburst) for modeling a direct IV input bypassing the connected controlled inputs
• Added: Changed default fontsize of axis labels from 10 to 12 points. The font size of axis titles is set to 120% of the axis label size
• Solved: The labels in (R-)plots from the code generator must always use the internal standard units
• Solved: The code generator ignored bioavailability in case of a TAbsorption object. The $fbio macro can now be used for this
• Solved: The AUC variable in TIntegrator is now not read-only in order to allow proper code generation
• Solved: Export code generator now also handles objects that use external symbols (dynamic references)
• Solved: A simulation with normal outputs also requires an initial fixed 4-step run-in phase just like when normal outputs are off
• Solved: Reset to population ([P]-button) should only reset parameters and not variables, constants and options
• Solved: Added deep protection against zero volumes (at Parm.Value level). Fixes issues with volumes connected to XRegression
• Solved: Check on zero parameter values when converting SD's between different parametrizations
• Solved: The 'Reset to population' engine function should set the individual SD to 0 and not to the population SD
• Solved: Do not apply shrinkage correction if Nobs <= Npk (Npk is number of parameters excluding weigh parameters)
• Solved: Create base FitAlgorithm object if model only contains common parameters (now not abstract class anymore)
• Solved: Pass-trough mode in PSystemic is now set using the new Pass option (Off or On). V=0 cannot be used anymore
• Solved: The PFactor is reset to 1 when looking at the fit settings, but it also never applied. Both issues are fixed
• Solved: Protect against zero Bh, Bw and Ccr during import of clinical data in Excel format
• Solved: NonComp error in case the concentration unit lacks a "/" (e.g. with shorthand molar units like µM)
• Solved: Moments class in numerical library (NumLib.dll) should check a variance of 0 before calculations (FP-Exception)
• Solved: Error with MultiSim simulation if object observations properties are visible in the designer
• Solved: Object with alternative parametrizations must update the parameters in the Reset event (or CalcNorm)
• Solved: With Excel import, the patient gender must be stored in the model (was ok) and in the MwPharm data segment (nok)
• Solved: with Excel import, the numeric identifiers for gender and race were not working
• Updated: Changing a population parameter now automatically updates the inidvidual parameter
• Updated: The population view is now the default view
• Updated: A covariate chain is now automatically repaired if one of the participating objects is deleted
• Updated: Added "reset to population values" button [P] to the toolbar
• Updated: Added covariate chain tracking using [C], [V] and [X] buttons
• Updated: Use an optional "EX" column ("1"=exclude) on the index/data sheet in order to exclude patients/observations