qbiocode.apps.quvine.baselines.adapters module#

Method Adapters for Baseline Methods

This module provides adapter functions that bridge between the registry’s config objects and the actual baseline method implementations. Each adapter: 1. Takes a config object and other required parameters 2. Converts config to method-specific parameters 3. Calls the actual baseline method 4. Returns the embedding in a consistent format

Summary#

Functions:

`get_adapter`	Get the adapter function for a method.
`run_appnp_adapter`	Adapter for APPNP baseline.
`run_baseline_filter_adapter`	Adapter for baseline filter methods (heat/poly without quantum calibration).
`run_baseline_gcnmf_adapter`	Adapter for baseline GCN-MF.
`run_gat_adapter`	Adapter for GAT-based methods.
`run_graphgps_adapter`	Adapter for GraphGPS-based methods.
`run_graphsage_adapter`	Adapter for GraphSAGE baseline.
`run_netmf_adapter`	Adapter for NetMF baseline.
`run_node2vec_adapter`	Adapter for Node2Vec baseline.
`run_quvine_gcnmf_adapter`	Adapter for quantum-calibrated GCN-MF.
`run_quvine_heat_adapter`	Adapter for quantum-calibrated heat kernel filter.
`run_quvine_poly_adapter`	Adapter for quantum-calibrated polynomial filter.
`run_quvine_sgns_adapter`	Adapter for the QuVINE SGNS walk embeddings (quvine_rwr/ctqw/dtqw).

Reference#

run_node2vec_adapter(graph_data, config, **kwargs)[source]#

Adapter for Node2Vec baseline.

Parameters:

graph_data – NetworkX graph
config (Node2VecConfig) – Node2VecConfig object
**kwargs – Additional arguments (ignored)

Return type:

ndarray

Returns:

Embedding matrix (n_nodes x dimensions)

run_appnp_adapter(graph_data, config, **kwargs)[source]#

Adapter for APPNP baseline.

Parameters:

graph_data – NetworkX graph
config (APPNPConfig) – APPNPConfig object
**kwargs – Additional arguments (ignored)

Return type:

ndarray

Returns:

Embedding matrix (n_nodes x dimensions)

run_graphsage_adapter(graph_data, config, **kwargs)[source]#

Adapter for GraphSAGE baseline.

Parameters:

graph_data – NetworkX graph
config (GraphSAGEConfig) – GraphSAGEConfig object
**kwargs – Additional arguments (ignored)

Return type:

ndarray

Returns:

Embedding matrix (n_nodes x dimensions)

run_netmf_adapter(graph_data, config, **kwargs)[source]#

Adapter for NetMF baseline.

Parameters:

graph_data – NetworkX graph
config (NetMFConfig) – NetMFConfig object
**kwargs – Additional arguments (ignored)

Return type:

ndarray

Returns:

Embedding matrix (n_nodes x dimensions)

run_baseline_filter_adapter(graph_data, config, **kwargs)[source]#

Adapter for baseline filter methods (heat/poly without quantum calibration).

Parameters:

graph_data – NetworkX graph
config (BaselineFilterConfig) – BaselineFilterConfig object
**kwargs – Additional arguments (ignored)

Return type:

ndarray

Returns:

Embedding matrix (n_nodes x embedding_dim)

run_quvine_heat_adapter(graph_data, config, q_targets, **kwargs)[source]#

Adapter for quantum-calibrated heat kernel filter.

Parameters:

graph_data – NetworkX graph
config (QuvineFilterConfig) – QuvineFilterConfig object
q_targets (List) – Quantum targets for calibration
**kwargs – Additional arguments (ignored)

Return type:

ndarray

Returns:

Embedding matrix (n_nodes x embedding_dim)

run_quvine_poly_adapter(graph_data, config, q_targets, **kwargs)[source]#

Adapter for quantum-calibrated polynomial filter.

Parameters:

graph_data – NetworkX graph
config (QuvineFilterConfig) – QuvineFilterConfig object
q_targets (List) – Quantum targets for calibration
**kwargs – Additional arguments (ignored)

Return type:

ndarray

Returns:

Embedding matrix (n_nodes x embedding_dim)

run_baseline_gcnmf_adapter(graph_data, config, **kwargs)[source]#

Adapter for baseline GCN-MF.

Parameters:

graph_data – NetworkX graph
config (GCNMFConfig) – GCNMFConfig object
**kwargs – Additional arguments (ignored)

Return type:

ndarray

Returns:

Embedding matrix (n_nodes x embedding_dim)

run_quvine_gcnmf_adapter(graph_data, config, q_targets, **kwargs)[source]#

Adapter for quantum-calibrated GCN-MF.

Parameters:

graph_data – NetworkX graph
config (QuvineGCNMFConfig) – QuvineGCNMFConfig object
q_targets (List) – Quantum targets for calibration
**kwargs – Additional arguments (ignored)

Return type:

ndarray

Returns:

Embedding matrix (n_nodes x embedding_dim)

run_quvine_sgns_adapter(graph_data, config, **kwargs)[source]#

Adapter for the QuVINE SGNS walk embeddings (quvine_rwr/ctqw/dtqw).

Runs the shared SGNS core (views -> walks -> corpus -> word2vec) for the single config.walk_kind and returns that walk kind’s embedding, with rows in list(graph_data.nodes) order.

Parameters:

graph_data – NetworkX graph
config (QuvineSGNSConfig) – QuvineSGNSConfig carrying the walk kind and full OmegaConf cfg
**kwargs – Additional arguments (ignored; SGNS needs no quantum targets)

Return type:

ndarray

Returns:

Embedding matrix (n_nodes x embedding_dim)

run_gat_adapter(graph_data, config, q_targets=None, **kwargs)[source]#

Adapter for GAT-based methods.

Handles all GAT variants: - raw: Standard GAT - heat_qcal_ctqw/dtqw/rwr: Quantum-calibrated with heat kernel - poly_qcal_ctqw/dtqw/rwr: Quantum-calibrated with polynomial filter

Parameters:

graph_data – NetworkX graph
config (GATMethodConfig) – GATMethodConfig object
q_targets (Optional[List]) – Quantum targets (optional, required for quantum variants)
**kwargs – Additional arguments (ignored)

Return type:

ndarray

Returns:

Embedding matrix (n_nodes x embedding_dim)

run_graphgps_adapter(graph_data, config, q_targets=None, **kwargs)[source]#

Adapter for GraphGPS-based methods.

Handles all GraphGPS variants: - raw: Standard GraphGPS - heat_qcal_ctqw/dtqw/rwr: Quantum-calibrated with heat kernel - poly_qcal_ctqw/dtqw/rwr: Quantum-calibrated with polynomial filter

Parameters:

graph_data – NetworkX graph
config (GraphGPSMethodConfig) – GraphGPSMethodConfig object
q_targets (Optional[List]) – Quantum targets (optional, required for quantum variants)
**kwargs – Additional arguments (ignored)

Return type:

ndarray

Returns:

Embedding matrix (n_nodes x embedding_dim)

get_adapter(method_name)[source]#

Get the adapter function for a method.

Parameters:: method_name (str) – Name of the method
Returns:: Adapter function
Raises:: KeyError – If method not found