Importing an existing network

This function will allow us to import a previously saved network. Again, note the return value that makes this a fluent interface:

public static Network ImportNetwork()
{

Get the filename of the previously saved network. Once opened, deserialize it into a network structure that we will deal with. (If it didn't work for some reason, abort!):

  var dn = GetHelperNetwork();
  if (dn == null) 
  return null;

Create a new Network and a list of neurons to populate:

var network = new Network();
  var allNeurons = new List<Neuron>();

Copy over the learning rate and the momentum that was previously saved:

network.LearningRate = dn.LearningRate;
  network.Momentum = dn.Momentum;

Create input layers from our imported network data:

foreach (var n in dn.InputLayer)
  {
    var neuron = new Neuron
    {
      Id = n.Id,
      Bias = n.Bias,
      BiasDelta = n.BiasDelta,
      Gradient = n.Gradient,
      Value = n.Value
    };

    network.InputLayer?.Add(neuron);
    allNeurons.Add(neuron);
  }

Create our Hidden Layers from our imported network data:

  foreach (var layer in dn.HiddenLayers)
  {
    var neurons = new List<Neuron>();
    foreach (var n in layer)
    {
      var neuron = new Neuron
      {
        Id = n.Id,
        Bias = n.Bias,
        BiasDelta = n.BiasDelta,
        Gradient = n.Gradient,
        Value = n.Value
      };

      neurons.Add(neuron);
      allNeurons.Add(neuron);
    }
    network.HiddenLayers?.Add(neurons);
  }

Create the OutputLayer neurons from our imported data:

foreach (var n in dn.OutputLayer)
  {
    var neuron = new Neuron
    {
      Id = n.Id,
      Bias = n.Bias,
      BiasDelta = n.BiasDelta,
      Gradient = n.Gradient,
      Value = n.Value
    };

    network.OutputLayer?.Add(neuron);
    allNeurons.Add(neuron);
  }

Finally, create the synapses that tie everything together:

  foreach (var syn in dn.Synapses)
  {
    var synapse = new Synapse{ Id = syn.Id };
    var inputNeuron = allNeurons.First(x =>x.Id==syn.InputNeuronId);
    var outputNeuron = allNeurons.First(x =>x.Id==syn.OutputNeuronId);
    synapse.InputNeuron = inputNeuron;
    synapse.OutputNeuron = outputNeuron;
    synapse.Weight = syn.Weight;
    synapse.WeightDelta = syn.WeightDelta;

    inputNeuron?.OutputSynapses?.Add(synapse);
    outputNeuron?.InputSynapses?.Add(synapse);
  }
  return network;
}

The following is where we manually enter the data to be used by the network:

public NNManager GetTrainingDataFromUser()
{
var numDataSets = GetInput("\tHow many datasets are you going to enter? ", 1, int.MaxValue);

  var newDatasets = new List<NNDataSet>();
  for (var i = 0; i<numDataSets; i++)
  {
    var values = GetInputData($"\tData Set {i + 1}: ");
    if (values == null)
    {
      return this;
    }

    var expectedResult = GetExpectedResult($"\tExpected Result for Data 
    Set {i + 1}: ");
    if (expectedResult == null)
    {
      return this;
    }

    newDatasets.Add(newNNDataSet(values, expectedResult));
  }

  _dataSets = newDatasets;
  return this;
}