FourParameterLogistic predict errors #19
Description
When using the FourParameterLogistic predict method, if a new y value is greater than y_max_parameter, an error will be thrown. This results in a root of a negative number and NumPy doesn't allow for complex numbers. This occurs regularly in immunoassays, so we need to create an exception to catch the error. Currently, np.nan is return, we need to decide to inform the user of the problem or if we want to use an absolute value and return a number with an error message.