It looks like Prime insists that you maximize for all the parameters of the function.
One workaround could be the introduction of a dummy function, as shown below. But I am not sure if the value for lamda is realistic and valid. At least its far away from the guess value but then it seems to fit the data.