|
5 | 5 | STATE_SIZE=5, |
6 | 6 | INPUT_SIZE=2, |
7 | 7 | ANGLE_IDX=2, # state vector index which is angle, used to fit diff in |
8 | | - R=np.diag([1.0e-06, 1.0e-06]), # control cost |
9 | | - Q=np.diag([10, 10, 10, 3, 0]), # state cost | x, y, theta, v, omega |
| 8 | + R=np.diag([1.0e-04, 1.0e-04]), # control cost |
| 9 | + Q=np.diag([30, 30, 30, 3, 0]), # state cost | x, y, theta, v, omega |
10 | 10 | Sf=np.diag([0, 0, 0, 0, 0]), # final state cost |
11 | 11 | ) |
12 | 12 |
|
@@ -70,7 +70,7 @@ class Config: |
70 | 70 | min_speed=80, |
71 | 71 | min_dist=0, # if agent is within this distance from trajectory end the goal is considered achieved |
72 | 72 | # ? for trajectories from data |
73 | | - px_to_cm=1 / 30.8, # convert px values to cm |
| 73 | + px_to_cm=1 / 8, # convert px values to cm |
74 | 74 | # dist_th=60, # keep frames only after moved away from start location |
75 | 75 | dist_th=-1, |
76 | 76 | resample=True, # if True when using tracking trajectory resamples it |
@@ -116,6 +116,11 @@ def __init__(self,): |
116 | 116 | self.Q = params["Q"] |
117 | 117 | self.Sf = params["Sf"] |
118 | 118 |
|
| 119 | + # Adjust mouse length for plotting |
| 120 | + self.mouse["length"] = ( |
| 121 | + self.mouse["length"] * self.trajectory["px_to_cm"] |
| 122 | + ) |
| 123 | + |
119 | 124 | def config_dict(self): |
120 | 125 | return dict( |
121 | 126 | dt=self.dt, |
|
0 commit comments