Functions — github.com/IBM/AssetOpsBench

CodeHub MCPcopy Create free account

hub / github.com/IBM/AssetOpsBench / functions

Functions947 in github.com/IBM/AssetOpsBench

⨍ Functions 947◇ Types & classes 171↳ Endpoints 13

↓ 2 callersFunctionrender_summary

(report: EvalReport)

src/evaluation/report.py:173

↓ 2 callersFunctionreset

Drop databases for a clean state. Returns the dropped names. Default: drop every user database (CouchDB GET /_all_dbs, system DBs excluded).

src/couchdb/init_data.py:111

↓ 2 callersMethodrun

Run one direct LLM call and return an AgentResult.

src/agent/direct_llm_agent/runner.py:80

↓ 2 callersMethodrun

Run the full plan-execute loop for a question. Steps: 1. Discover available servers from registered MCP servers. 2. Use t

src/agent/plan_execute/runner.py:110

↓ 2 callersFunctionscenario

src/servers/wo/tests/test_workorders.py:84

↓ 2 callersFunctionscenario

src/servers/wo/tests/test_models_boundary.py:23

↓ 2 callersMethodsummary

Return a compact summary (no raw data).

src/servers/vibration/data_store.py:54

↓ 1 callersFunction_RMSE

(y_true, y_pred, axis=None)

src/servers/tsfm/metrics.py:11

↓ 1 callersMethod__init__

(self, inner: LLMBackend)

src/agent/plan_execute/runner.py:38

↓ 1 callersFunction_absolute_error

src/servers/tsfm/anomaly.py:20

↓ 1 callersFunction_accel_g_to_velocity_rms_mms

Convert an acceleration signal (in g) to RMS velocity (in mm/s) within the ISO 10816 band (default 10-1000 Hz). Method: frequency-domain

src/servers/vibration/main.py:85

↓ 1 callersFunction_aggregate_score_summary

Aggregate static_json-style score.details across all results. Per-scenario key-level details stay in each result. Here we summarize the numer

src/evaluation/report.py:29

↓ 1 callersFunction_amp_loss

(outputs, targets)

src/servers/tsfm/metrics.py:124

↓ 1 callersMethod_annotate_span

( self, span, trajectory: Trajectory, answer: str, started: float )

src/agent/stirrup_agent/runner.py:185

↓ 1 callersFunction_ashift_loss

(outputs, targets)

src/servers/tsfm/metrics.py:144

↓ 1 callersMethod_build_client

Build a Stirrup LLM client for the configured model id.

src/agent/stirrup_agent/runner.py:86

↓ 1 callersMethod_build_code_provider

Build the sandboxed code-execution provider for the code track.

src/agent/stirrup_agent/runner.py:121

↓ 1 callersFunction_build_llm

src/servers/fmsr/main.py:98

↓ 1 callersFunction_build_llm

(model_id: str)

src/agent/cli.py:60

↓ 1 callersFunction_build_llm

(model_id: str)

src/agent/direct_llm_agent/cli.py:45

↓ 1 callersMethod_build_mcp_provider

Build a Stirrup ``MCPToolProvider`` for the AssetOpsBench servers. Each server is a stdio process launched exactly as the other runners

src/agent/stirrup_agent/runner.py:102

↓ 1 callersFunction_build_parser

src/benchmark/scenario_suite_runner.py:515

↓ 1 callersFunction_build_parser

src/evaluation/cli.py:15

↓ 1 callersMethod_build_tools

src/agent/stirrup_agent/runner.py:135

↓ 1 callersFunction_bulk_insert

(db, docs, batch_size=500)

src/couchdb/loader.py:255

↓ 1 callersMethod_call

src/mcphub/__init__.py:80

↓ 1 callersFunction_call_asset2fm

Query the LLM for failure modes of an asset. Retries up to _MAX_RETRIES times. Results are cached to avoid redundant LLM calls for the same asset.

src/servers/fmsr/main.py:139

↓ 1 callersFunction_call_tool

Connect to an MCP server via stdio and call a tool.

src/agent/plan_execute/executor.py:317

↓ 1 callersFunction_candidate_part

Find the message/tool part inside common OpenCode event shapes.

src/agent/opencode_agent/runner.py:252

↓ 1 callersFunction_clean_final_answer

Remove <think> blocks and keep only the final structured answer.

src/agent/direct_llm_agent/runner.py:40

↓ 1 callersFunction_coerce

(col, val, int_f, float_f, json_f)

src/couchdb/loader.py:68

↓ 1 callersFunction_coerce_tool_input

src/agent/opencode_agent/runner.py:270

↓ 1 callersFunction_collect_docs

Resolve a manifest source ("default"/path/dir/list/inline docs) to parsed docs. Relative data paths resolve against the scenario folder (``base_d

src/couchdb/loader.py:120

↓ 1 callersFunction_conformal_set

(y_pred, score_threshold, nonconformity_score="absolute_error")

src/servers/tsfm/anomaly.py:38

↓ 1 callersFunction_couchdb_reachable

src/servers/iot/tests/conftest.py:13

↓ 1 callersFunction_create_indexes

src/couchdb/loader.py:245

↓ 1 callersFunction_df_dt_stats

(pd_dataset, date_col="Timestamp", intervals_dic=None)

src/servers/tsfm/dataquality.py:66

↓ 1 callersFunction_df_single_columns_condition

(df, condition_dic=None)

src/servers/tsfm/dataquality.py:97

↓ 1 callersFunction_dq_timeseries_segmentation

( pd_merge, filtering_params=None, timestamp_tag="Timestamp" )

src/servers/tsfm/dataquality.py:253

↓ 1 callersFunction_efficient_nan_removal

(pd_table, preference_tie="row")

src/servers/tsfm/dataquality.py:114

↓ 1 callersFunction_encode_otlp_json

Encode *spans* as a single-line OTLP-JSON ``TracesData`` message. Isolated for testability; also keeps the `_internal` import contained to on

src/observability/file_exporter.py:59

↓ 1 callersFunction_ensure_db

src/couchdb/loader.py:217

↓ 1 callersFunction_extract_balanced_structure

Extract first balanced {...}, [...], or (...) from noisy text.

src/evaluation/scorers/static_json.py:107

↓ 1 callersFunction_extract_content

Extract text from MCP tool call result content.

src/agent/plan_execute/executor.py:330

↓ 1 callersFunction_final_answer

Return text from the final assistant message only.

src/agent/opencode_agent/runner.py:406

↓ 1 callersFunction_find_largest_tsfm_checkpoint_directory

(root_dir: str)

src/servers/tsfm/forecasting.py:645

↓ 1 callersFunction_find_repo_root

Locate AssetOpsBench root whether this file is at root or in a suite folder.

src/servers/vibration/tests/test_mcp_e2e.py:52

↓ 1 callersFunction_finetune_ttm_hf

( df_dataframe, dataset_config_dictionary, model_config, save_model_dir, n_finetune, n

src/servers/tsfm/forecasting.py:376

↓ 1 callersFunction_from_plan_execute

(steps: list[Any], model: str)

src/evaluation/metrics.py:131

↓ 1 callersFunction_from_sdk_trajectory

(traj: dict, model: str)

src/evaluation/metrics.py:38

↓ 1 callersFunction_get_tsad_aligned_data

( df_data, dataset_config, ad_config, tsmodel_prediction_dictionary )

src/servers/tsfm/anomaly.py:368

↓ 1 callersFunction_get_tsfm_dataloaders

( df_dataframe, model_config, dataset_config_dictionary, scaling=False )

src/servers/tsfm/anomaly.py:296

↓ 1 callersFunction_install_design

(db, design_doc)

src/couchdb/loader.py:226

↓ 1 callersFunction_install_fake_openai

Install a stub ``openai`` module that records call kwargs.

src/llm/tests/test_backends.py:13

↓ 1 callersFunction_is_answer_text

Whether a part contributes to the user-facing answer.

src/agent/opencode_agent/runner.py:372

↓ 1 callersFunction_is_step_finish

True for OpenCode step-finish boundaries.

src/agent/opencode_agent/runner.py:398

↓ 1 callersFunction_iter_manifest_paths

Return path-like strings from a scenario manifest value.

src/benchmark/scenario_suite_runner.py:166

↓ 1 callersFunction_kurtosis

Excess kurtosis (Fisher definition, normal = 0).

src/servers/vibration/data_store.py:21

↓ 1 callersFunction_load_scenario_dir

Load scenario folders from a directory containing groundtruth.txt files. Supports folder names like: scenario_34/ groundtruth.t

src/evaluation/loader.py:78

↓ 1 callersFunction_load_scenario_file

src/evaluation/loader.py:126

↓ 1 callersFunction_make_id

(key, cfg, doc)

src/couchdb/loader.py:173

↓ 1 callersFunction_make_json_compatible

Recursively convert an object to a JSON-serializable form.

src/servers/tsfm/io.py:55

↓ 1 callersMethod_match

(self, doc, selector)

src/servers/wo/tests/test_workorders.py:46

↓ 1 callersFunction_maybe_install_judge

(judge_model: str | None)

src/evaluation/cli.py:69

↓ 1 callersFunction_merge_text

Combine repeated emissions for one text part id. OpenCode streams can look like snapshots (new value contains the previous value) or deltas (

src/agent/opencode_agent/runner.py:383

↓ 1 callersFunction_nest

(doc, dotted_key, value)

src/couchdb/loader.py:78

↓ 1 callersFunction_normalise

(doc, key, cfg, transform)

src/couchdb/loader.py:183

↓ 1 callersFunction_opencode_usage_tokens

Return input/output totals from OpenCode's step-finish token schema.

src/agent/opencode_agent/runner.py:309

↓ 1 callersFunction_parse_arguments

Stirrup ToolCall.arguments is a JSON string; parse defensively.

src/agent/stirrup_agent/trajectory.py:69

↓ 1 callersFunction_parse_file

src/couchdb/loader.py:116

↓ 1 callersFunction_parse_numbered_list

Parse a numbered list response into a plain list of strings.

src/servers/fmsr/main.py:68

↓ 1 callersFunction_parse_relevancy

Parse a 3-line relevancy response into {answer, reason, temporal_behavior}.

src/servers/fmsr/main.py:78

↓ 1 callersFunction_parse_review

src/evaluation/scorers/llm_judge.py:155

↓ 1 callersFunction_phase_loss

(outputs, targets)

src/servers/tsfm/metrics.py:153

↓ 1 callersMethod_register_default_workflows

src/mcphub/__init__.py:251

↓ 1 callersFunction_remove_df_rows_by_single_column_condition

(df, column_name, condition)

src/servers/tsfm/dataquality.py:158

↓ 1 callersFunction_reset_provider

Swap in a fresh global TracerProvider for the next test.

src/observability/tests/test_file_exporter.py:18

↓ 1 callersFunction_resolve

src/couchdb/loader.py:130

↓ 1 callersMethod_resolve

(self, name: str)

src/mcphub/__init__.py:275

↓ 1 callersMethod_resolve

src/evaluation/evaluator.py:79

↓ 1 callersFunction_resolve_manifest

Return (manifest, base_dir). scenario_id None → the default manifest. Otherwise the scenario's own manifest (folder form scenario_<id>/manife

src/couchdb/init_data.py:71

↓ 1 callersFunction_resolve_manifest_data_path

Resolve a manifest data path using the same search order as init_data.

src/benchmark/scenario_suite_runner.py:183

↓ 1 callersMethod_run

src/mcphub/__init__.py:94

↓ 1 callersMethod_score_one

( self, scenario: Scenario, traj: PersistedTrajectory )

src/evaluation/evaluator.py:56

↓ 1 callersFunction_serialize_trajectory

Reduce a trajectory to plain JSON-friendly Python. Handles the two shapes in this repo: SDK runners' ``Trajectory`` dataclass, and plan-execu

src/observability/persistence.py:89

↓ 1 callersFunction_server_params

Build StdioServerParameters for a given server module. Uses sys.executable (env-agnostic), strips LLM credentials (air-gap), and disables OTE

src/servers/vibration/tests/test_mcp_e2e.py:92

↓ 1 callersFunction_strip_markdown_fence

Strip markdown code fences if present.

src/evaluation/scorers/static_json.py:92

↓ 1 callersFunction_strip_think_blocks

Remove model-private <think>...</think> blocks before scoring.

src/evaluation/evaluator.py:108

↓ 1 callersFunction_threshold_condition_function

(threshold, condition_type="<")

src/servers/tsfm/dataquality.py:16

↓ 1 callersFunction_time_series_frequency_interval_segmentation

( df, time_column, lower_bound=14, upper_bound=16 )

src/servers/tsfm/dataquality.py:168

↓ 1 callersFunction_time_series_segment_quality_summary

(df, timestamp_column, segments_column)

src/servers/tsfm/dataquality.py:228

↓ 1 callersFunction_tracing_enabled

src/observability/tracing.py:46

↓ 1 callersFunction_trajectory_to_text

Flatten a trajectory to a text blob for the LLM-As-Judge prompt.

src/evaluation/evaluator.py:98

↓ 1 callersFunction_transform_for

Optional per-collection transform: a function named <key> in transforms.py.

src/couchdb/loader.py:161

↓ 1 callersFunction_tsfm_available

src/servers/tsfm/tests/conftest.py:12

↓ 1 callersFunction_tsfm_dataloader_to_array

( dataset_calibration, ix_target_features, x_context_window=-1 )

src/servers/tsfm/anomaly.py:331

↓ 1 callersFunction_usage_from_raw_events

Extract token usage from raw event schemas kept for parser fallback.

src/evaluation/metrics.py:93

↓ 1 callersMethod_validate_judge_model

(self, scorer_name: str, traj: PersistedTrajectory)

src/evaluation/evaluator.py:82

↓ 1 callersFunction_validate_scorer_default

src/evaluation/cli.py:81

↓ 1 callersFunction_validate_time_series_segments

( df_segment, segment_tag="segment_id", timestamp_tag="Timestamp", p_nan_rows=1, p_nan_col

src/servers/tsfm/dataquality.py:185

← previous next →201–300 of 947, ranked by callers