add NaN check for reward manager. start vel penalty

Project/Assets/ML-Agents/Examples/SharedAssets/Scripts/JointDriveController.cs

         public void SetJointStrength(float strength)
         {
             var rawVal = (strength + 1f) * 0.5f * thisJdController.maxJointForceLimit;
-            var jd = new JointDrive
-            {
-                positionSpring = thisJdController.maxJointSpring,
-                positionDamper = thisJdController.jointDampen,
-                maximumForce = rawVal
-            };
+//            var remappedVal = Mathf.InverseLerp(1, 1, strength); 
+            JointDrive jd = joint.slerpDrive;
+            jd.positionSpring = thisJdController.maxJointSpring;
+            jd.positionDamper = thisJdController.jointDampen;
+            jd.maximumForce = rawVal;
+//            var jd = new JointDrive
+//            {
+//                positionSpring = thisJdController.maxJointSpring,
+//                positionDamper = thisJdController.jointDampen,
+//                maximumForce = rawVal
+//            };
+//        public void SetJointStrength(float strength)
+//        {
+//            var rawVal = (strength + 1f) * 0.5f * thisJdController.maxJointForceLimit;
+//            var jd = new JointDrive
+//            {
+//                positionSpring = thisJdController.maxJointSpring,
+//                positionDamper = thisJdController.jointDampen,
+//                maximumForce = rawVal
+//            };
+//            joint.slerpDrive = jd;
+//            currentStrength = jd.maximumForce;
+//        }
     }
 
     public class JointDriveController : MonoBehaviour
 
         public float jointDampen;
+        [Min(1)]
-        [HideInInspector] public List<BodyPart> bodyPartsList = new List<BodyPart>();
+        public List<BodyPart> bodyPartsList = new List<BodyPart>();
+//        [HideInInspector] public List<BodyPart> bodyPartsList = new List<BodyPart>();
         const float k_MaxAngularVelocity = 50.0f;
 
         /// <summary>

Project/Assets/ML-Agents/Examples/SharedAssets/Scripts/RewardManager.cs

         public float cumulativeThisEpisode;
         public float cumulativeThisSession;
 
-        public float maxRewardThisSession;
+//        public float maxRewardThisSession;
+
+        public int lastNaNStep;
 //        public Reward(string k)
 //        public Reward()
 //        {
     {
         rewardsDict[key].rawVal = rawVal;
         float scaledVal = rawVal * rewardsDict[key].rewardScalar;
-        rewardsDict[key].maxRewardThisSession = scaledVal * maxSteps;
+
+        //if we get a NaN, set the step
+        if (float.IsNaN(scaledVal))
+            rewardsDict[key].lastNaNStep = m_thisAgent.StepCount;
+        
+//        rewardsDict[key].maxRewardThisSession = scaledVal * maxSteps;
+        
         m_thisAgent.AddReward(scaledVal);
     }
 

Project/Assets/ML-Agents/Examples/Walker/Scripts/WalkerAgent.cs

     public float headHeightOverFeetReward; //reward for standing up straight-ish
     public float hurryUpReward = -1; //don't waste time
     public RewardManager rewardManager;
+    public float bpVelPenaltyThisStep = 0;
     void UpdateRewards()
     {
         var cubeForward = orientationCube.transform.forward;
         rewardManager.UpdateReward("matchSpeed", matchSpeedReward);
         rewardManager.UpdateReward("lookAtTarget", lookAtTargetReward);
         rewardManager.UpdateReward("headHeightOverFeet", headHeightOverFeetReward);
-        rewardManager.UpdateReward("hurryUp", hurryUpReward/MaxStep); 
+        rewardManager.UpdateReward("hurryUp", hurryUpReward/MaxStep);
+
+        bpVelPenaltyThisStep = 0;
+        foreach (var item in m_JdController.bodyPartsList)
+        {
+            bpVelPenaltyThisStep += item.rb.velocity.magnitude;
+        }
     }
     
 //    void FixedUpdate()