Error codes

This is the error codes list returned by int StraightforwardNeuralNetwork::isValid() metohd.

Code	Error Name	Description
Neural Network error
101	invalid number of inputs	Number of inputs must be between 0 and 2,073,600 (1080 x 1920).
102	invalid number of layers	Number of layer must be between 0 and 1,000.
Layer error
201	invalid number of neurons	Neurons by layer must be between 1 and 1,000,000.
202	invalid number of outputs	Number of outputs must be equal to number of neurons inside layer.
203	invalid weight neurons	neurons created with wrong number of weights
Neuron error
301	wrong bias	Bias must be equal to 1.
302	wrong number of weights	Number of weights must be between 1 and 1,000,000.
303	wrong weight	Weights must be stay between -10,000 and +10,000.
304	wrong weight	Error during the creation of a recurrent neuron.
Data error
401	wrong value	A value after normalization is not in range [-1;1].
402	null set	A composite have a null set.
403	bad data	Data is badly created.
404	bad temporal data	Temporal data is badly created.
405	wrong size	Error in size of data.
406	wrong number of labels	Data type classification must have at least 2 labels.
Optimizer
501	invalid learning rate	Learning rate must be between 0 and 1.
502	invalid momentum	Momentum must be between 0 and 1.
503	invalid beta1	Momentum must be between 0 and 1.
504	invalid beta2	Momentum must be between 0 and 1.
505	invalid epsilon	Momentum must be between 0 and 1.

isValid method don't throw error because the method is always supposed to be valid, otherwise it is an error in the library code.