Make initial guess handle a numpy array of points for walker starting points

[TroyGustke]

When this feature is used after implementation, the initial guess should be in the shape (n_walkers, n_parameters)

[AdityaSavara]

o During plotting, if initial guess has multiple walkers (multi dimensional) then switch to self.UserInput.mu_prior
 This requires checking if the initial guess is a string or not.
• IN the init area, where we do the check for strings, make a comment noting that if we ever support direct multidimensional arrays for initial guesses (for multiple walkers) that we will need to change the checks in the plotting function at that time also.
• WE should just (at this time) make a helper function called “checkInitialGuessType”, and that function should return the initial guess type (string, blank [blank is zero length], single parameter set, multiple walkers parameter set, other [not supported])
 For example the below line needs to be changed:
• self.mu_guess_SimulatedOutput = simulationFunction( self.UserInput.InputParameterInitialGuess) #Do NOT use self.UserInput.model['InputParameterInitialGuess'] because that won't work with reduced parameter space requests.
 This needs to be done for every plot type that plots the initial guess (including heatmap)

[AdityaSavara]

Related card:
#285