1 reporting.wandb.runs

reporting.wandb.runs

Discover W&B runs and project their metadata into comparison tables.

1.1 Classes

Name	Description
WandbApi	Minimal W&B API interface used by these utilities.
WandbRun	Minimal W&B Run interface used by these utilities.

1.1.1 WandbApi

reporting.wandb.runs.WandbApi()

Minimal W&B API interface used by these utilities.

1.1.1.1 Methods

Name	Description
viewer	Return metadata for the authenticated viewer and its teams.
projects	Iterate projects visible under one W&B entity.
runs	Iterate runs under an `entity/project` path in API order.
run	Fetch one run by its fully qualified W&B path.

1.1.1.1.1 viewer

reporting.wandb.runs.WandbApi.viewer()

Return metadata for the authenticated viewer and its teams.

1.1.1.1.2 projects

reporting.wandb.runs.WandbApi.projects(entity)

Iterate projects visible under one W&B entity.

1.1.1.1.3 runs

reporting.wandb.runs.WandbApi.runs(path, order=None)

Iterate runs under an entity/project path in API order.

1.1.1.1.4 run

reporting.wandb.runs.WandbApi.run(path)

Fetch one run by its fully qualified W&B path.

1.1.2 WandbRun

reporting.wandb.runs.WandbRun()

Minimal W&B Run interface used by these utilities.

1.1.2.1 Attributes

Name	Description
id	Stable W&B run identifier.
name	Human-readable run display name.
state	Remote lifecycle state such as `running` or `finished`.
group	Optional experiment group label.
job_type	Optional W&B job-role label.
tags	Optional ordered run tags.
created_at	Creation timestamp in the representation returned by the W&B client.
summary	Final or latest scalar summary payload.
config	Serialized run configuration payload.

1.1.2.2 Methods

Name	Description
history	Return history data (typically a pandas.DataFrame or list of dicts).
scan_history	Iterate every matching history row without W&B downsampling.

1.1.2.2.1 history

reporting.wandb.runs.WandbRun.history(keys=None, samples=None)

Return history data (typically a pandas.DataFrame or list of dicts).

1.1.2.2.2 scan_history

reporting.wandb.runs.WandbRun.scan_history(keys=None)

Iterate every matching history row without W&B downsampling.

1.2 Functions

Name	Description
build_autoresearch_run_dataframe	Return the read-only autoresearch evidence rows embedded in W&B configs.
build_run_dataframes	Build meta/summary/config dataframes from W&B runs.

1.2.1 build_autoresearch_run_dataframe

reporting.wandb.runs.build_autoresearch_run_dataframe(runs)

Return the read-only autoresearch evidence rows embedded in W&B configs.

The bridge writes this compact identity data under aria_autoresearch; metrics remain ordinary W&B summaries. Runs without that namespace are intentionally excluded, so the panel never infers campaign state from a run name, tag, or mutable W&B history.

1.2.2 build_run_dataframes

reporting.wandb.runs.build_run_dataframes(runs)

Build meta/summary/config dataframes from W&B runs.

Returns new dataframes suitable for method chaining.

# reporting.wandb.runs { #aria_nbv.reporting.wandb.runs } `reporting.wandb.runs` Discover W&B runs and project their metadata into comparison tables. ## Classes | Name | Description | | --- | --- | | [WandbApi](#aria_nbv.reporting.wandb.runs.WandbApi) | Minimal W&B API interface used by these utilities. | | [WandbRun](#aria_nbv.reporting.wandb.runs.WandbRun) | Minimal W&B Run interface used by these utilities. | ### WandbApi { #aria_nbv.reporting.wandb.runs.WandbApi } ```python reporting.wandb.runs.WandbApi() ``` Minimal W&B API interface used by these utilities. #### Methods | Name | Description | | --- | --- | | [viewer](#aria_nbv.reporting.wandb.runs.WandbApi.viewer) | Return metadata for the authenticated viewer and its teams. | | [projects](#aria_nbv.reporting.wandb.runs.WandbApi.projects) | Iterate projects visible under one W&B entity. | | [runs](#aria_nbv.reporting.wandb.runs.WandbApi.runs) | Iterate runs under an `entity/project` path in API order. | | [run](#aria_nbv.reporting.wandb.runs.WandbApi.run) | Fetch one run by its fully qualified W&B path. | ##### viewer { #aria_nbv.reporting.wandb.runs.WandbApi.viewer } ```python reporting.wandb.runs.WandbApi.viewer() ``` Return metadata for the authenticated viewer and its teams. ##### projects { #aria_nbv.reporting.wandb.runs.WandbApi.projects } ```python reporting.wandb.runs.WandbApi.projects(entity) ``` Iterate projects visible under one W&B entity. ##### runs { #aria_nbv.reporting.wandb.runs.WandbApi.runs } ```python reporting.wandb.runs.WandbApi.runs(path, order=None) ``` Iterate runs under an `entity/project` path in API order. ##### run { #aria_nbv.reporting.wandb.runs.WandbApi.run } ```python reporting.wandb.runs.WandbApi.run(path) ``` Fetch one run by its fully qualified W&B path. ### WandbRun { #aria_nbv.reporting.wandb.runs.WandbRun } ```python reporting.wandb.runs.WandbRun() ``` Minimal W&B Run interface used by these utilities. #### Attributes | Name | Description | | --- | --- | | [id](#aria_nbv.reporting.wandb.runs.WandbRun.id) | Stable W&B run identifier. | | [name](#aria_nbv.reporting.wandb.runs.WandbRun.name) | Human-readable run display name. | | [state](#aria_nbv.reporting.wandb.runs.WandbRun.state) | Remote lifecycle state such as `running` or `finished`. | | [group](#aria_nbv.reporting.wandb.runs.WandbRun.group) | Optional experiment group label. | | [job_type](#aria_nbv.reporting.wandb.runs.WandbRun.job_type) | Optional W&B job-role label. | | [tags](#aria_nbv.reporting.wandb.runs.WandbRun.tags) | Optional ordered run tags. | | [created_at](#aria_nbv.reporting.wandb.runs.WandbRun.created_at) | Creation timestamp in the representation returned by the W&B client. | | [summary](#aria_nbv.reporting.wandb.runs.WandbRun.summary) | Final or latest scalar summary payload. | | [config](#aria_nbv.reporting.wandb.runs.WandbRun.config) | Serialized run configuration payload. | #### Methods | Name | Description | | --- | --- | | [history](#aria_nbv.reporting.wandb.runs.WandbRun.history) | Return history data (typically a pandas.DataFrame or list of dicts). | | [scan_history](#aria_nbv.reporting.wandb.runs.WandbRun.scan_history) | Iterate every matching history row without W&B downsampling. | ##### history { #aria_nbv.reporting.wandb.runs.WandbRun.history } ```python reporting.wandb.runs.WandbRun.history(keys=None, samples=None) ``` Return history data (typically a pandas.DataFrame or list of dicts). ##### scan_history { #aria_nbv.reporting.wandb.runs.WandbRun.scan_history } ```python reporting.wandb.runs.WandbRun.scan_history(keys=None) ``` Iterate every matching history row without W&B downsampling. ## Functions | Name | Description | | --- | --- | | [build_autoresearch_run_dataframe](#aria_nbv.reporting.wandb.runs.build_autoresearch_run_dataframe) | Return the read-only autoresearch evidence rows embedded in W&B configs. | | [build_run_dataframes](#aria_nbv.reporting.wandb.runs.build_run_dataframes) | Build meta/summary/config dataframes from W&B runs. | ### build_autoresearch_run_dataframe { #aria_nbv.reporting.wandb.runs.build_autoresearch_run_dataframe } ```python reporting.wandb.runs.build_autoresearch_run_dataframe(runs) ``` Return the read-only autoresearch evidence rows embedded in W&B configs. The bridge writes this compact identity data under ``aria_autoresearch``; metrics remain ordinary W&B summaries. Runs without that namespace are intentionally excluded, so the panel never infers campaign state from a run name, tag, or mutable W&B history. ### build_run_dataframes { #aria_nbv.reporting.wandb.runs.build_run_dataframes } ```python reporting.wandb.runs.build_run_dataframes(runs) ``` Build meta/summary/config dataframes from W&B runs. Returns new dataframes suitable for method chaining.