Heuristic

Hi,

i'm a beginner at mlagents, i played around with the example environments for a while, and took on to make my own flappy bird ai (i thought it'll be easier considering almost everyone has done this, and i'll be able to find a lot of resources in case i get stuck)

So i'm stuck here at configuring the actions,

can't control the agent properly, the button press is not registered everytime,

and since i can't get the heuristic right, i suppose i won't be able to train the AI properly.

Heuristic Function

Code (CSharp):

1.     public override float[] Heuristic()
2.     {
3.         var action = new float[1];      
4.         if (Input.GetKeyDown(KeyCode.Space))
5.         {
6.             Debug.Log("Pressed");
7.             action[0] = 1 ;
8.         }
9.         else
10.         {
11.             action[0] = 0;
12.         }
13.         return action;
14.     }

OnActionReceived Function

Code (CSharp):

1.     public override void OnActionReceived(float[] vectorAction)
2.     {
3.         {
4.             AddReward(0.01f);
5.             int action = Mathf.FloorToInt(vectorAction[0]);
6.             if(action == 0)
7.             {
8.             }
9.             if(action == 1)
10.             {
11.                 m_AgentRb.velocity = Vector3;
12. m_AgentRb.AddForce(Vector2.up * tapForce, ForceMode2D.Force);
13.             }
14.         }
15.     }

can't figure out what's wrong with this code

Please Help

 

try this Out :

Code (CSharp):

1. public override void Heuristic(float[] actionsOut)
2.     {
3.         actionsOut[0] = Input.GetKeyDown(KeyCode.Space) ? 1 : 0 ;
4.     }
5.  

and make sure your behavior type in behavior parameters is set to default for training Section and Heuristic for debugging Section