dsh-eval-guard 0.7.4 → 0.7.5

package/assets/neural-pendulum/rl-feasibility/reference_controller.py

@@ -1,112 +0,0 @@
-"""Non-RL feasibility check: a pole-placement stabiliser around upright.
-
-This is NOT a learned policy and NOT part of the RL results.  It answers a
-single question: is the upright hold physically reachable with the fixed
-torque mapping (tau1, -tau3, tau3) and |u| <= 1?
-
-The controller uses privileged full state (theta and the simulator omega),
-so it is an upper bound / existence proof only.  It is deliberately simple:
-    u = clip(-K [wrap(theta-pi); omega], -1, 1)
-with K from pole placement on the linearisation at upright.
-
-Writes results/reference_controller.json.
-"""
-from __future__ import annotations
-
-import json
-from pathlib import Path
-
-import numpy as np
-from scipy.signal import place_poles
-
-from sim import EPISODE_STEPS, N_ACT, Params, PendulumSim, make_torch_consts
-from task import EVAL_OMEGA0, EVAL_THETA0, NumpyTracker, sample_train_init, wrap_to_pi
-
-ROOT = Path(__file__).resolve().parent
-
-
-def design(p: Params, poles=(-5.0, -6.0, -7.0, -8.0, -9.0, -10.0)):
-    M = p.A
-    K = np.diag(p.g * p.gvec)
-    Minv = np.linalg.inv(M)
-    B = np.zeros((3, 2))
-    B[0, 0] = 1.0
-    B[1, 1] = -1.0
-    B[2, 1] = 1.0
-    Be = Minv @ B
-    A = np.zeros((6, 6))
-    A[:3, 3:] = np.eye(3)
-    A[3:, :3] = Minv @ K
-    B6 = np.vstack([np.zeros((3, 2)), Be])
-    res = place_poles(A, B6, np.asarray(poles, dtype=float))
-    return res.gain_matrix
-
-
-def rollout(p, gain, theta0, omega0, episode_steps=EPISODE_STEPS,
-            alpha=0.5):
-    sim = PendulumSim(p, episode_steps=episode_steps)
-    tr = NumpyTracker()
-    theta = sim.reset(theta0, omega0)
-    tr.reset()
-    sat = 0
-    uf = np.zeros(2)
-    for _ in range(episode_steps):
-        x = np.r_[wrap_to_pi(theta - np.pi), sim._omega]
-        raw = np.clip(-gain @ x, -1.0, 1.0)
-        # first-order low-pass on the torque command: the unfiltered
-        # pole-placement controller chatters at 100 Hz and only holds ~1 s.
-        uf = alpha * raw + (1.0 - alpha) * uf
-        sat += int(np.abs(uf).max() > 1.0)
-        theta = sim.step(uf)
-        tr.update(theta, 0.0)
-    summ = tr.summary()
-    return dict(t_can=float(summ["t_can"]),
-                ever_up=float(summ["ever_up"]),
-                final_ang_err=float(np.abs(wrap_to_pi(theta - np.pi)).max()),
-                sat_frac=float(sat / episode_steps),
-                final_theta=theta.tolist())
-
-
-def main():
-    p = Params()
-    gain = design(p)
-    # This is a LOCAL stabiliser: start near upright (theta = pi + offset).
-    fixed_theta0 = np.pi + EVAL_THETA0
-    fixed = rollout(p, gain, fixed_theta0, EVAL_OMEGA0)
-    rng = np.random.default_rng(0)
-    th = np.pi + rng.uniform(-0.05, 0.05, size=(32, 3))
-    om = np.zeros((32, 3))
-    rows = [rollout(p, gain, th[i], om[i]) for i in range(th.shape[0])]
-    tc = np.array([r["t_can"] for r in rows])
-    out = dict(
-        note=("Non-RL existence check.  Pole-placement stabiliser using "
-              "privileged full state (theta and simulator omega); not a "
-              "learned policy.  Shows whether the upright hold is physically "
-              "reachable with the fixed torque mapping.  The torque command is "
-              "low-pass filtered (alpha=0.5); the unfiltered controller "
-              "chatters and only holds ~1 s."),
-        poles=[-5.0, -6.0, -7.0, -8.0, -9.0, -10.0],
-        action_filter_alpha=0.5,
-        gain=gain.tolist(),
-        fixed_start=dict(theta0=fixed_theta0.tolist(),
-                         offset_from_upright=EVAL_THETA0.tolist(),
-                         omega0=EVAL_OMEGA0.tolist(), **fixed),
-        random_starts=dict(n=len(rows), success_rate=float((tc >= 10.0).mean()),
-                           mean_t_can=float(tc.mean()),
-                           median_t_can=float(np.median(tc)),
-                           max_t_can=float(tc.max()),
-                           rows=rows),
-    )
-    out_path = ROOT / "results" / "reference_controller.json"
-    with open(out_path, "w") as f:
-        json.dump(out, f, indent=2)
-    print(json.dumps(dict(fixed_t_can=fixed["t_can"],
-                          fixed_final_err=fixed["final_ang_err"],
-                          fixed_sat=fixed["sat_frac"],
-                          random_success=out["random_starts"]["success_rate"],
-                          random_mean_t_can=out["random_starts"]["mean_t_can"]),
-                     indent=2))
-
-
-if __name__ == "__main__":
-    main()

package/assets/neural-pendulum/rl-feasibility/results/reference_controller.json

@@ -1,417 +0,0 @@
-{
-  "note": "Non-RL existence check.  Pole-placement stabiliser using privileged full state (theta and simulator omega); not a learned policy.  Shows whether the upright hold is physically reachable with the fixed torque mapping.  The torque command is low-pass filtered (alpha=0.5); the unfiltered controller chatters and only holds ~1 s.",
-  "poles": [
-    -5.0,
-    -6.0,
-    -7.0,
-    -8.0,
-    -9.0,
-    -10.0
-  ],
-  "action_filter_alpha": 0.5,
-  "gain": [
-    [
-      -17.82008272259022,
-      -90.53020345591291,
-      -37.60973879998583,
-      -17.574950427476214,
-      -21.006026390694455,
-      -9.992697352896354
-    ],
-    [
-      3.4221790257027664,
-      6.7210733064834685,
-      6.823263521680083,
-      1.9296331848889634,
-      2.183847052498074,
-      1.386819759832599
-    ]
-  ],
-  "fixed_start": {
-    "theta0": [
-      3.154592653589793,
-      3.133592653589793,
-      3.162592653589793
-    ],
-    "offset_from_upright": [
-      0.013,
-      -0.008,
-      0.021
-    ],
-    "omega0": [
-      0.0,
-      0.0,
-      0.0
-    ],
-    "t_can": 29.99000000000189,
-    "ever_up": 1.0,
-    "final_ang_err": 2.1760371282653068e-14,
-    "sat_frac": 0.0,
-    "final_theta": [
-      3.141592653589815,
-      3.141592653589791,
-      3.141592653589787
-    ]
-  },
-  "random_starts": {
-    "n": 32,
-    "success_rate": 1.0,
-    "mean_t_can": 24.79125000000196,
-    "median_t_can": 27.840000000001897,
-    "max_t_can": 29.99000000000189,
-    "rows": [
-      {
-        "t_can": 29.99000000000189,
-        "ever_up": 1.0,
-        "final_ang_err": 8.43769498715119e-15,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535898016,
-          3.1415926535897905,
-          3.1415926535897936
-        ]
-      },
-      {
-        "t_can": 18.890000000002082,
-        "ever_up": 1.0,
-        "final_ang_err": 5.595524044110789e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.141592653589849,
-          3.1415926535897984,
-          3.141592653589774
-        ]
-      },
-      {
-        "t_can": 18.0500000000021,
-        "ever_up": 1.0,
-        "final_ang_err": 5.3290705182007514e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535898464,
-          3.1415926535897984,
-          3.141592653589776
-        ]
-      },
-      {
-        "t_can": 11.180000000001748,
-        "ever_up": 1.0,
-        "final_ang_err": 5.717648576819556e-12,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535840755,
-          3.1415926535880225,
-          3.141592653591231
-        ]
-      },
-      {
-        "t_can": 29.99000000000189,
-        "ever_up": 1.0,
-        "final_ang_err": 4.707345624410664e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.14159265358984,
-          3.141592653589798,
-          3.1415926535897785
-        ]
-      },
-      {
-        "t_can": 21.580000000002023,
-        "ever_up": 1.0,
-        "final_ang_err": 2.708944180085382e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.141592653589766,
-          3.141592653589795,
-          3.141592653589803
-        ]
-      },
-      {
-        "t_can": 28.81000000000189,
-        "ever_up": 1.0,
-        "final_ang_err": 1.4210854715202004e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.141592653589779,
-          3.141592653589796,
-          3.1415926535897962
-        ]
-      },
-      {
-        "t_can": 29.99000000000189,
-        "ever_up": 1.0,
-        "final_ang_err": 4.085620730620576e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.141592653589834,
-          3.1415926535897913,
-          3.1415926535897762
-        ]
-      },
-      {
-        "t_can": 29.99000000000189,
-        "ever_up": 1.0,
-        "final_ang_err": 2.5757174171303632e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535897674,
-          3.141592653589791,
-          3.141592653589812
-        ]
-      },
-      {
-        "t_can": 16.970000000002123,
-        "ever_up": 1.0,
-        "final_ang_err": 3.572708751065079e-08,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926893168806,
-          3.1415926726829873,
-          3.1415926548413156
-        ]
-      },
-      {
-        "t_can": 14.390000000002178,
-        "ever_up": 1.0,
-        "final_ang_err": 5.062616992290714e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535898437,
-          3.141592653589805,
-          3.141592653589778
-        ]
-      },
-      {
-        "t_can": 29.99000000000189,
-        "ever_up": 1.0,
-        "final_ang_err": 3.952393967665557e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535897536,
-          3.141592653589795,
-          3.141592653589809
-        ]
-      },
-      {
-        "t_can": 24.62000000000196,
-        "ever_up": 1.0,
-        "final_ang_err": 3.952393967665557e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535898326,
-          3.1415926535897913,
-          3.1415926535897767
-        ]
-      },
-      {
-        "t_can": 29.99000000000189,
-        "ever_up": 1.0,
-        "final_ang_err": 4.218847493575595e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535898353,
-          3.1415926535897922,
-          3.141592653589776
-        ]
-      },
-      {
-        "t_can": 29.99000000000189,
-        "ever_up": 1.0,
-        "final_ang_err": 3.907985046680551e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.141592653589832,
-          3.1415926535897922,
-          3.141592653589779
-        ]
-      },
-      {
-        "t_can": 17.260000000002115,
-        "ever_up": 1.0,
-        "final_ang_err": 3.79389852867007e-11,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.141592653592213,
-          3.141592653561257,
-          3.141592653551854
-        ]
-      },
-      {
-        "t_can": 25.760000000001938,
-        "ever_up": 1.0,
-        "final_ang_err": 1.5987211554602254e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.141592653589809,
-          3.1415926535897936,
-          3.1415926535897833
-        ]
-      },
-      {
-        "t_can": 27.260000000001906,
-        "ever_up": 1.0,
-        "final_ang_err": 2.6201263381153694e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.141592653589767,
-          3.141592653589795,
-          3.141592653589802
-        ]
-      },
-      {
-        "t_can": 14.380000000002179,
-        "ever_up": 1.0,
-        "final_ang_err": 7.682743330406083e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535897163,
-          3.1415926535897825,
-          3.141592653589821
-        ]
-      },
-      {
-        "t_can": 24.21000000000197,
-        "ever_up": 1.0,
-        "final_ang_err": 4.574118861455645e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.141592653589839,
-          3.1415926535897922,
-          3.141592653589775
-        ]
-      },
-      {
-        "t_can": 25.350000000001945,
-        "ever_up": 1.0,
-        "final_ang_err": 3.3306690738754696e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.14159265358976,
-          3.1415926535897953,
-          3.141592653589806
-        ]
-      },
-      {
-        "t_can": 19.280000000002076,
-        "ever_up": 1.0,
-        "final_ang_err": 1.865174681370263e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535898118,
-          3.14159265358979,
-          3.141592653589787
-        ]
-      },
-      {
-        "t_can": 29.99000000000189,
-        "ever_up": 1.0,
-        "final_ang_err": 3.5083047578154947e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.141592653589758,
-          3.1415926535897953,
-          3.1415926535898073
-        ]
-      },
-      {
-        "t_can": 29.89000000000189,
-        "ever_up": 1.0,
-        "final_ang_err": 1.3322676295501878e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535898064,
-          3.14159265358979,
-          3.1415926535897896
-        ]
-      },
-      {
-        "t_can": 28.420000000001888,
-        "ever_up": 1.0,
-        "final_ang_err": 2.353672812205332e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535897696,
-          3.1415926535897962,
-          3.1415926535898024
-        ]
-      },
-      {
-        "t_can": 29.94000000000189,
-        "ever_up": 1.0,
-        "final_ang_err": 1.8207657603852567e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535898113,
-          3.1415926535897905,
-          3.1415926535897873
-        ]
-      },
-      {
-        "t_can": 29.53000000000189,
-        "ever_up": 1.0,
-        "final_ang_err": 4.263256414560601e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535898357,
-          3.1415926535897913,
-          3.1415926535897754
-        ]
-      },
-      {
-        "t_can": 29.99000000000189,
-        "ever_up": 1.0,
-        "final_ang_err": 4.4853010194856324e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535897483,
-          3.141592653589794,
-          3.1415926535898113
-        ]
-      },
-      {
-        "t_can": 29.88000000000189,
-        "ever_up": 1.0,
-        "final_ang_err": 2.6201263381153694e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535898193,
-          3.141592653589792,
-          3.1415926535897847
-        ]
-      },
-      {
-        "t_can": 21.020000000002035,
-        "ever_up": 1.0,
-        "final_ang_err": 4.1744385725905886e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.141592653589835,
-          3.1415926535897936,
-          3.141592653589775
-        ]
-      },
-      {
-        "t_can": 17.920000000002105,
-        "ever_up": 1.0,
-        "final_ang_err": 3.998623654410949e-11,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.141592653549807,
-          3.141592653569871,
-          3.1415926535902368
-        ]
-      },
-      {
-        "t_can": 28.82000000000189,
-        "ever_up": 1.0,
-        "final_ang_err": 3.419486915845482e-14,
-        "sat_frac": 0.0,
-        "final_theta": [
-          3.1415926535898273,
-          3.1415926535897913,
-          3.14159265358978
-        ]
-      }
-    ]
-  }
-}