AxoBench API
AxoBench API
Public AxoBench functions, classes, constants, and command-line options.
Package surface
This reference covers AxoBench 0.2.1 at revision 496e0c1a2dfd. Model evaluation, paired comparison, and diagnostics are exported lazily from axobench.benchmark; metrics, relevance, data readers, defect injection, and NeuronIO conversion are imported from their modules. The package needs NumPy and imports no model framework.
from axobench.benchmark import evaluate_model, evaluate_model_against_cache
from axobench.metrics import compute_core_metric_set
from axobench.data import ShardedNeuronIODataset
Exports of axobench.benchmark
| Name | Implementation | Reference |
|---|---|---|
BenchmarkCostModel |
axobench.benchmark.profiles.BenchmarkCostModel |
Source |
BenchmarkProfile |
axobench.benchmark.profiles.BenchmarkProfile |
Source |
BenchmarkSuite |
axobench.benchmark.suite.BenchmarkSuite |
Source |
DIAGNOSTIC_SUITES |
axobench.benchmark.dataset_schema.DIAGNOSTIC_SUITES |
Open |
DatasetSelector |
axobench.benchmark.selectors.DatasetSelector |
Open |
DiagnosticAuditSources |
axobench.benchmark.diagnostic_audits.DiagnosticAuditSources |
Source |
DiagnosticContext |
axobench.benchmark.diagnostic_rows.DiagnosticContext |
Open |
EvaluationConfig |
axobench.benchmark.iteration_evaluator.EvaluationConfig |
Open |
INTERVENTION_CONDITIONS |
axobench.benchmark.dataset_schema.INTERVENTION_CONDITIONS |
Open |
LOCAL_CORENEURON_HAY_COST_MODEL |
axobench.benchmark.profiles.LOCAL_CORENEURON_HAY_COST_MODEL |
Source |
ModelEvaluationCache |
axobench.benchmark.iteration_evaluator.ModelEvaluationCache |
Open |
MorphologyTransferEvaluator |
axobench.benchmark.morphology_transfer.MorphologyTransferEvaluator |
Source |
ORDINARY_SPLITS |
axobench.benchmark.dataset_schema.ORDINARY_SPLITS |
Open |
PairedInterventionPredictions |
axobench.benchmark.iteration_evaluator.PairedInterventionPredictions |
Open |
PopulationDynamicsConfig |
axobench.benchmark.population_dynamics.PopulationDynamicsConfig |
Source |
V1DatasetLayout |
axobench.benchmark.selectors.V1DatasetLayout |
Open |
V1_MORPHOLOGY_COUNT |
axobench.benchmark.dataset_schema.V1_MORPHOLOGY_COUNT |
Open |
adapt_branch_count |
axobench.benchmark.branch_adapter.adapt_branch_count |
Source |
adapter_fairness_diagnostic_rows |
axobench.benchmark.diagnostic_audits.adapter_fairness_diagnostic_rows |
Source |
analyze_population_perturbation |
axobench.benchmark.population_dynamics.analyze_population_perturbation |
Source |
analyze_population_spikes |
axobench.benchmark.population_dynamics.analyze_population_spikes |
Source |
build_axis_plan |
axobench.benchmark.profiles.build_axis_plan |
Source |
build_diagnostic_suite_catalog |
axobench.benchmark.dataset_schema.build_diagnostic_suite_catalog |
Open |
build_neuron_state_masks |
axobench.benchmark.state_metrics.build_neuron_state_masks |
Source |
build_v1_dataset_manifest_template |
axobench.benchmark.dataset_schema.build_v1_dataset_manifest_template |
Open |
compare_predictions |
axobench.benchmark.iteration_evaluator.compare_predictions |
Open |
compute_paired_event_intervention_metrics |
axobench.benchmark.structured_intervention.compute_paired_event_intervention_metrics |
Source |
compute_perturbation_stability |
axobench.benchmark.perturbation_stability.compute_perturbation_stability |
Source |
compute_regime_stratified_metrics |
axobench.benchmark.regime_stratified.compute_regime_stratified_metrics |
Source |
compute_state_conditioned_metrics |
axobench.benchmark.state_metrics.compute_state_conditioned_metrics |
Source |
compute_trace_shape_metrics |
axobench.benchmark.trace_shape.compute_trace_shape_metrics |
Source |
core_metric_comparison_rows |
axobench.benchmark.reports.core_metric_comparison_rows |
Source |
create_model_evaluation_cache |
axobench.benchmark.iteration_evaluator.create_model_evaluation_cache |
Open |
default_diagnostic_specs |
axobench.benchmark.dataset_schema.default_diagnostic_specs |
Source |
default_iteration_sample_limits |
axobench.benchmark.iteration_evaluator.default_iteration_sample_limits |
Open |
default_target_views |
axobench.benchmark.dataset_schema.default_target_views |
Open |
diagnostic_audit_rows |
axobench.benchmark.diagnostic_audits.diagnostic_audit_rows |
Source |
diagnostic_audit_rows_from_sources |
axobench.benchmark.diagnostic_audits.diagnostic_audit_rows_from_sources |
Source |
error_anatomy |
axobench.benchmark.error_anatomy.error_anatomy |
Open |
estimate_profile_cost |
axobench.benchmark.profiles.estimate_profile_cost |
Source |
evaluate_model |
axobench.benchmark.iteration_evaluator.evaluate_model |
Open |
evaluate_model_against_cache |
axobench.benchmark.iteration_evaluator.evaluate_model_against_cache |
Open |
evaluate_model_diagnostic_rows |
axobench.benchmark.diagnostic_rows.evaluate_model_diagnostic_rows |
Open |
evaluate_predictions |
axobench.benchmark.iteration_evaluator.evaluate_predictions |
Open |
evaluate_with_branch_adaptation |
axobench.benchmark.branch_adapter.evaluate_with_branch_adaptation |
Source |
format_evaluation_report |
axobench.benchmark.iteration_evaluator.format_evaluation_report |
Open |
format_model_comparison |
axobench.benchmark.iteration_evaluator.format_model_comparison |
Open |
get_benchmark_profile |
axobench.benchmark.profiles.get_benchmark_profile |
Source |
infer_pair_metadata |
axobench.benchmark.bundle.infer_pair_metadata |
Source |
list_benchmark_profiles |
axobench.benchmark.profiles.list_benchmark_profiles |
Source |
load_artifact_rows |
axobench.benchmark.diagnostic_audits.load_artifact_rows |
Source |
load_available_morphologies |
axobench.benchmark.morphology_transfer.load_available_morphologies |
Source |
load_intervention_pair_npz |
axobench.benchmark.runner.load_intervention_pair_npz |
Open |
load_model_evaluation_cache |
axobench.benchmark.iteration_evaluator.load_model_evaluation_cache |
Open |
load_selector_batch |
axobench.benchmark.runner.load_selector_batch |
Source |
make_event_dropout_inputs |
axobench.benchmark.structured_intervention.make_event_dropout_inputs |
Open |
make_selective_event_dropout_inputs |
axobench.benchmark.structured_intervention.make_selective_event_dropout_inputs |
Open |
make_site_silence_inputs |
axobench.benchmark.structured_intervention.make_site_silence_inputs |
Open |
make_structured_intervention_inputs |
axobench.benchmark.structured_intervention.make_structured_intervention_inputs |
Open |
make_temporal_jitter_inputs |
axobench.benchmark.structured_intervention.make_temporal_jitter_inputs |
Open |
metric_decoupling_diagnostic_rows |
axobench.benchmark.diagnostic_audits.metric_decoupling_diagnostic_rows |
Source |
morphology_contrast_diagnostic_rows |
axobench.benchmark.diagnostic_rows.morphology_contrast_diagnostic_rows |
Open |
morphology_routing_qc_rows |
axobench.benchmark.diagnostic_rows.morphology_routing_qc_rows |
Open |
multistep_branching_ratio |
axobench.benchmark.population_dynamics.multistep_branching_ratio |
Source |
paired_intervention_diagnostic_rows |
axobench.benchmark.diagnostic_rows.paired_intervention_diagnostic_rows |
Open |
protocol_coverage_diagnostic_rows |
axobench.benchmark.diagnostic_audits.protocol_coverage_diagnostic_rows |
Source |
provisional_v1_morphologies |
axobench.benchmark.dataset_schema.provisional_v1_morphologies |
Source |
reweight_to_rate_distribution |
axobench.benchmark.error_anatomy.reweight_to_rate_distribution |
Open |
robustness_directionality_diagnostic_rows |
axobench.benchmark.diagnostic_audits.robustness_directionality_diagnostic_rows |
Source |
run_model_diagnostics_on_selector |
axobench.benchmark.runner.run_model_diagnostics_on_selector |
Open |
run_paired_intervention_on_npz |
axobench.benchmark.runner.run_paired_intervention_on_npz |
Open |
run_paired_intervention_on_selector |
axobench.benchmark.runner.run_paired_intervention_on_selector |
Open |
run_teacher_qc_on_selector |
axobench.benchmark.runner.run_teacher_qc_on_selector |
Open |
run_v1_benchmark_bundle |
axobench.benchmark.bundle.run_v1_benchmark_bundle |
Open |
summarize_diagnostic_rows |
axobench.benchmark.reports.summarize_diagnostic_rows |
Open |
target_view_audit_diagnostic_rows |
axobench.benchmark.diagnostic_audits.target_view_audit_diagnostic_rows |
Source |
teacher_qc_diagnostic_rows |
axobench.benchmark.diagnostic_rows.teacher_qc_diagnostic_rows |
Open |
v1_morphologies |
axobench.benchmark.dataset_schema.v1_morphologies |
Open |
validate_diagnostic_rows |
axobench.benchmark.diagnostic_rows.validate_diagnostic_rows |
Open |
validate_v1_manifest_template |
axobench.benchmark.dataset_schema.validate_v1_manifest_template |
Open |
write_core_metric_comparison_table |
axobench.benchmark.reports.write_core_metric_comparison_table |
Source |
write_diagnostic_rows |
axobench.benchmark.diagnostic_rows.write_diagnostic_rows |
Open |
write_diagnostic_summary |
axobench.benchmark.reports.write_diagnostic_summary |
Open |
write_evaluation_report |
axobench.benchmark.iteration_evaluator.write_evaluation_report |
Open |
write_model_comparison |
axobench.benchmark.iteration_evaluator.write_model_comparison |
Open |
write_model_evaluation_cache |
axobench.benchmark.iteration_evaluator.write_model_evaluation_cache |
Open |
Exports marked Source link to their module. They include the combined BenchmarkSuite, benchmark profiles and cost models, branch adapters, morphology transfer, population dynamics, state, regime, trace-shape, and perturbation metrics, artifact audits, and comparison tables. They are part of the package but outside the evaluation protocol documented in this reference.
Modules
- Model evaluation
- Paired comparison
- Core metric functions
- Spike event metrics
- IRon relevance
- Trace datasets
- Dataset layout and schema
- Intervention inputs
- NeuronIO conversion
- Defect injection
- Diagnostic rows
- Error anatomy
- axobench CLI
Source
Every source link is pinned to revision 496e0c1a2dfd. Repository access is currently required to open the source links. The generator reads signatures and defaults with AST parsing and the CLI options from the argparse parser.