This solves a problem with the ordering of parameters for ABC. Previously, it was necessary to specify the model parameters before any initial conditions, and the initial conditions needed to be specified in the order they appeared in the model.
The parameters can now be specified in any order. From this, the appropriate order is determined and stored in self.par_order, which is then used to re-order the sampled parameters immediately before they are fed into the model.
This solves a problem with the ordering of parameters for ABC. Previously, it was necessary to specify the model parameters before any initial conditions, and the initial conditions needed to be specified in the order they appeared in the model.
The parameters can now be specified in any order. From this, the appropriate order is determined and stored in self.par_order, which is then used to re-order the sampled parameters immediately before they are fed into the model.