Datasets:
benchflow
/
frontierphysics-pr32-evidence

Dataset card Data Studio Files
xet
 Community
text
stringclasses
308 values
"maximum_quadratic_design_condition",
"median_first_order_rms_residual",
"median_second_order_rms_residual",
"median_quadratic_signal_rms",
"median_residual_reduction_fraction",
"fraction_residual_reduction_above_threshold",
):
assert math.isclose(
float(summary[key]),
float(expected[key]),
rel_tol=2e-9,
abs_tol=2e-11,
)
for key in (
"residue_count",
"directed_edge_count",
"minimum_neighbor_count",
"maximum_neighbor_count",
"minimum_quadratic_design_rank",
):
assert int(summary[key]) == int(expected[key])
> assert summary["top_nonaffine_residue_keys"] == expected_top
E AssertionError: assert ['S001-R00362...-R00340', ...] == ['S024-R00004...-R00390', ...]
E
E At index 0 diff: 'S001-R00362' != 'S024-R00004'
E
E Full diff:
E [
E - 'S024-R00004',
E - 'S024-R00024',...
E
E ...Full output truncated (64 lines hidden), use '-vv' to show
test_outputs.py:1150: AssertionError
_________ test_submitted_executable_handles_hidden_neighborhood_radius _________
def test_submitted_executable_handles_hidden_neighborhood_radius() -> None:
data = input_data()
> artifacts, hidden_reference, hidden_nonaffine = run_submitted_executable(
data,
HIDDEN_RADIUS,
"hidden-radius",
)
test_outputs.py:1839:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
test_outputs.py:1575: in run_submitted_executable
validate_nonaffine_summary_against_input(
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
artifacts = {'arrays': {'compression': array([0.63796991, 0.71070637, 0.70676394, ..., 0.61725815, 0.52854174,
0.23303206])... 0.8790912591451675, 'input_schema_version': 'neutral-protein-transition-v1', 'maximum_neighbor_count': 183, ...}, ...}
fixture = {'residue_key': array(['S000-R00001', 'S000-R00002', 'S000-R00003', ..., 'S025-R00141',
'S025-R00142', 'S025-R0...S025', 'S025', 'S025'], dtype='<U4'), 'segment_position': array([ 1, 2, 3, ..., 141, 142, 143], dtype=int32), ...}
reference = {'deformation_hessian': array([[[[ 0.01521155, -0.01241159, 0.00527232],
[-0.01241159, 0.04924336, -0.01364..._order_rms_residual': array([2.39071866, 2.5620272 , 2.82955642, ..., 4.39353368, 3.34198284,
2.30062038]), ...}
def validate_nonaffine_summary_against_input(
artifacts: dict[str, object],
fixture: dict[str, np.ndarray | str],
reference: dict[str, np.ndarray | float | int],
) -> None:
arrays = artifacts["nonaffine_arrays"]
summary = artifacts["nonaffine_summary"]
residue_key = np.asarray(fixture["residue_key"])
count = residue_key.size
quadratic_signal = np.asarray(arrays["quadratic_signal_rms"])
residual_reduction = np.asarray(
arrays["residual_reduction_fraction"]
)
ordering = sorted(
range(count),
key=lambda index: (
-float(quadratic_signal[index]),
str(residue_key[index]),
),
)
expected_top = [
str(residue_key[index])
for index in ordering[: min(TOP_K, count)]
]
expected = {
"model_radius_angstrom": float(reference["model_radius_angstrom"]),
"model_radius_multiplier": NONAFFINE_RADIUS_MULTIPLIER,
"nonaffine_reduction_threshold": NONAFFINE_REDUCTION_THRESHOLD,
"residue_count": count,
"directed_edge_count": int(reference["directed_edge_count"]),
"minimum_neighbor_count": int(
np.min(arrays["model_neighbor_count"])
),
"maximum_neighbor_count": int(
np.max(arrays["model_neighbor_count"])
),
"minimum_quadratic_design_rank": int(
np.min(arrays["quadratic_design_rank"])
),
"maximum_quadratic_design_condition": float(
np.max(arrays["quadratic_design_condition"])
),