public override ActionResponse GetActionResponse(int environmentSymbol, float[] state, float reward, bool done)
{
ActionResponse response = new ActionResponse();
// Send message with (symbol, *state, reward, done)
network.SetFloatInBuffer(environmentSymbol, 0);
network.SetArrayFloatsInBuffer(state, 1);
network.SetFloatInBuffer(reward, observationDimension + 1);
network.SetFloatInBuffer(done ? 1f : 0f, observationDimension + 2);
network.Send();
// Receive status symbol & action
network.ReceiveFloatsAndString(1 + actionDimension, receiveBuffer, ref receivedMessage);
AgentActionSymbol agentSymbol = (AgentActionSymbol)Mathf.RoundToInt(receiveBuffer[0]);
for (int i = 1; i < receiveBuffer.Length; i++)
{
actions[i - 1] = receiveBuffer[i];
}
// Return
response.symbol = agentSymbol;
response.actions = actions;
response.msg = receivedMessage;
return(response);
}
// Constructor to use for training, i.e. when Python starts everything
public PythonAgentController(Environment environment, int portSend, int portReceive)
{
this.environment = environment;
network = new UDPNetwork(portSend, portReceive);
observationDimension = environment.ObservationDimension();
actionDimension = environment.ActionDimension();
network.AllocateSendBuffer(1 + 4 + 4 * observationDimension);
/* Send network msg to Python (12 bytes)
* 1. port (int = 4 bytes) (port that python sends and unity receives)
* 2. observation dimension (int = 4 bytes)
* 3. action dimension (int = 4 bytes)
*/
byte[] obsdimBytes = BitConverter.GetBytes(observationDimension);
byte[] actdimBytes = BitConverter.GetBytes(actionDimension);
byte[] introMsg = new byte[2 * 4];
introMsg[0] = obsdimBytes[0];
introMsg[1] = obsdimBytes[1];
introMsg[2] = obsdimBytes[2];
introMsg[3] = obsdimBytes[3];
introMsg[4] = actdimBytes[0];
introMsg[5] = actdimBytes[1];
introMsg[6] = actdimBytes[2];
introMsg[7] = actdimBytes[3];
network.Send(introMsg);
}