var can be a single variable name or list of variable names. In the case of an exception (a problem with the integration), r

Currently only accepts equally-spaced times. var should be a single string or list of strings of variable names To get time

Return the expected average integrated error between the model producing the fittingData and the fittingModel with its curren

Generates an ensemble of parameter sets using ensemble generator self.ensGen, and uses each of these as a starting point for

Minimize the cost of a model using Levenberg-Marquardt.

Minimize the cost of a model using Levenberg-Marquardt in terms of log parameters.

Creates data structure on disk for keeping track of fitting over increasing amounts of data and multiple conditions. After initializ

Returns model output as array: output[0] = time series for species 1, output[1] = time series for species 2, ...

Passes on kwargs to networkList2DOT. (The networkList is hard-coded, so this function always returns the same network.

Passes on kwargs to networkList2DOT.

Passes on kwargs to networkList2DOT. indepParamMax (None) : List of maximum values for indep params weightScale (1.)

Make one network figure for each condition. See documentation for FittingProblem.PowerLawFittingProblem.networkFigureBestModel.

Adds Gaussian noise to data: mean 0, stdev noiseFracSize*("typical value" of variable) (By default, the "typical value" is the m

Sum of all ensemble times plus max of minimization times over ALL models tested. (Time taken when running in parallel.)

matrix plot, Number of measurements vs. model number index (0) : Index of fittingProblem to plot fpdList (None)

plotDivisibleBy (None) : plot only N_Ds divisible by plotDivisibleBy percent (50.) : confidence interval size (0 t

Plot showing how well you're doing at fitting the function that takes current values to current derivatives. (I think assume

Plots mean over fpds in fpdList.

Plots mean performance versus mean number of parameters, along with the resulting Pareto front. plotDivisibleBy (None) :

indices (None) : If a list of indepParamsList indices, plots only these indices. Otherwise plots

See documentation for FittingProblem.plotResults.

Set all 'started' work to 'unstarted'. (Leave 'finished' alone.)

Each worker node runs this function to look for and perform work. endTime (None) : Stop work if endTime hours (wall time)

Sets parameters to random values given by the function randFunc (by default, uniformly distributed on [0,1) ).

Sets parameters to random values given by the function randFunc (by default, uniformly distributed on [0,1) ).

Overwrite stopFittingN values with given value. resetFitAllDone (True) : If True, set all fitAllDone to False.

Splits an allFpdsDict (eg for bestOutOfSampleCorrs with returnErrors=True)

Test 1 + 1 = 2

Test basic MPI functionality

Test that fitAll produces reasonable results on an easy test problem.

Test that a basic fitting problem can be initialized correctly

Test generateEnsemble_parallel

Test localFitToData_parallel

Test that a basic SloppyCell fitting model can be initialized correctly

Sum of all ensemble times over ALL models tested.

Sum of all minimization times over ALL models tested.

Sum of all minimization times over ALL models tested.

Total number of times daeint was called. Sum of (cost calls) * (number of measurements) (grad calls + ensemble steps) * (1

Sum of all cost calls plus grad calls over ALL models tested.

Sum of all ensemble times plus minimization times over ALL models tested. (Time taken when running in series.)

Returns the typical ranges of initial conditions for the parameters in self.indepParamNames. Useful when creating random ini

Generates a BNGL file (BioNetGen input file) for the PhosphorylationFit problem. n : number of phosphorylation sites

Writes a BNGL file that directly simulates a given .net file.

Changes parameter values by directly modifying the original .net file. Assumes the original .net file has all the parameters set to 1.0