ibl_alignment_gui.plugins.channel_prediction

Functions

`_compute_region_id_predictions`	Run a region-id `predict` callable and convert its output to region boundaries.
`_load_inference_model`	Load inference model via GUI dialog; reveal its region options and refresh on success.
`_load_spatial_model`	Load the spatial model via dialog; reveal its region option and refresh on success.
`_on_data_loaded`	Data-load hook: reset per-session plugin state and expose available model region options.
`_plot_region_panels`	Compute (and cache) a shank's prediction for `model` and draw it.
`_refresh_model_options`	Add the region-plot options for models that should currently be available.
`_set_local_features`	Prompt for a per-channel features parquet and use it for inference.
`compute_cumulative_predictions`	Cumulative prediction model using the inference model.
`compute_inference_predictions`	Prediction model using the inference model.
`compute_mapping_predictions`	Return brain regions based on a specified atlas mapping, as an example prediction model.
`compute_spatial_encoder_predictions`	Prediction model using the spatial encoder.
`get_region_boundaries`	Get the boundaries of brain regions along the histology track.
`setup`	Register the Channel Prediction plugin and (when available) its menu.

Classes

ChannelPrediction

Plugin loader that computes and plots per-channel region predictions.

class ibl_alignment_gui.plugins.channel_prediction.ChannelPrediction(controller)[source] 

Bases: object

Plugin loader that computes and plots per-channel region predictions.

Registered as the Channel Prediction plugin’s loader; plot_regions() dispatches the selected region model to the matching compute_* function and draws it on each shank.

__init__(controller)[source] 

Store the controller and cache its brain atlas.

Parameters:: controller (AlignmentGUIController) – The main application controller.

plot_regions(model, data_only=True)[source] 

Compute and plot the selected region model across all shanks.

Looks up model in the dispatch map and runs the matching compute_* function on each shank.

Parameters:

model (str) – Region-model key (e.g. ‘Beryl’, ‘Cosmos’, ‘Spatial Encoder’, ‘Inference Model’, ‘Inference Cumulative’). Unknown keys are ignored.
data_only (bool) – Reserved for signature compatibility with the region-plot callback; not used here.

Return type:

None

ibl_alignment_gui.plugins.channel_prediction.compute_cumulative_predictions(controller, items)[source] 

Cumulative prediction model using the inference model.

Parameters:

controller (AlignmentGUIController) – The main application controller.
items (ShankController) – The shank controller containing the model and view for the current shank.

Returns:

A bunch containing the probability of predicted brain regions along the probe, or None if no prediction is available.

Return type:

Bunch or None

ibl_alignment_gui.plugins.channel_prediction.compute_inference_predictions(controller, items)[source] 

Prediction model using the inference model.

Parameters:

controller (AlignmentGUIController) – The main application controller.
items (ShankController) – The shank controller containing the model and view for the current shank.

Returns:

The predicted brain regions along the probe, or None if no prediction is available.

Return type:

Bunch or None

ibl_alignment_gui.plugins.channel_prediction.compute_mapping_predictions(controller, items, mapping='Beryl')[source] 

Return brain regions based on a specified atlas mapping, as an example prediction model.

Parameters:

controller ('AlignmentGUIController') – The main application controller.
items ('ShankController') – The shank controller containing the model and view for the current shank.
mapping (str) – The atlas mapping to use for predictions (e.g., ‘Beryl’ or ‘Cosmos’).

Returns:

A bunch containing the predicted brain regions.

Return type:

Bunch

ibl_alignment_gui.plugins.channel_prediction.compute_spatial_encoder_predictions(controller, items)[source] 

Prediction model using the spatial encoder.

Parameters:

controller (AlignmentGUIController) – The main application controller.
items (ShankController) – The shank controller containing the model and view for the current shank.

Returns:

The predicted brain regions along the probe, or None if no prediction is available.

Return type:

Bunch or None

ibl_alignment_gui.plugins.channel_prediction.get_region_boundaries(regions, depths)[source] 

Get the boundaries of brain regions along the histology track.

Parameters:

regions (dict) – The brain regions along the histology track.
depths (np.ndarray) – The depths along the histology track, in meters.

Returns:

A bunch containing the region boundaries, labels, and colours.

Return type:

Bunch

ibl_alignment_gui.plugins.channel_prediction.setup(controller)[source] 

Register the Channel Prediction plugin and (when available) its menu.

Always installs the plugin state and its ChannelPrediction loader. When ephysatlas is importable, also adds the “Channel Prediction” menu (load inference/spatial models, load a features file) and registers a data-loaded callback that exposes the model region options.

Parameters:: controller (AlignmentGUIController) – The main application controller.
Return type:: None