Dashboard¶

SCLabDashboard¶

SCLabDashboard ¶

SCLabDashboard(adata_or_filepath_or_url: AnnData | str | None = None, name: str = 'SCLab Dashboard', counts_layer: str = 'counts', batch_key: str | None = None, copy: bool = False)

Bases: GridBox

Interactive single-cell RNA-seq analysis dashboard for Jupyter notebooks.

Integrates data loading, preprocessing, clustering, visualization, and analysis into a single tabbed widget. Built on top of AnnData/Scanpy with an event-driven architecture for reactive updates between components.

Parameters:

Name	Type	Description	Default
`adata_or_filepath_or_url`	`AnnData or str or None`	Input data as an AnnData object, a file path (.h5, .h5ad, or 10x MTX folder), or a URL to a remote .h5 or .h5ad file. If None, a data loader widget is displayed instead.	`None`
`name`	`str`	Display name for the dashboard. Default is "SCLab Dashboard".	`'SCLab Dashboard'`
`counts_layer`	`str`	Name of the layer to use (and create if absent) for raw counts. Default is "counts".	`'counts'`
`batch_key`	`str or None`	Key in `adata.obs` identifying batches/samples for batch-aware steps such as Integration and Neighbors. Default is None.	`None`
`copy`	`bool`	If True, work on a copy of the input AnnData. Default is False.	`False`

Attributes:

Name	Type	Description
`ds`	`SCLabDataset`	Shorthand property for the dataset component.
`pl`	`Plotter`	Shorthand property for the plotter component.
`pr`	`Processor`	Shorthand property for the processor component.

Examples:

>>> import scanpy as sc
>>> from sclab import SCLabDashboard
>>> adata = sc.datasets.pbmc68k_reduced()
>>> dashboard = SCLabDashboard(adata, name="PBMC Analysis")
>>> display(dashboard)

SCLabDataset¶

SCLabDataset ¶

SCLabDataset(adata: AnnData, name: str = 'SCLabDataset', counts_layer: str = 'counts', copy: bool = True, broker: EventBroker | None = None)

Bases: EventClient

Wrapper around AnnData with interactive tables and event-driven updates.

Manages observation and variable metadata, row selection, and data representations (obsm entries). Publishes events via the :class:~sclab.event.EventBroker whenever the data or selection changes so that the :class:~sclab.dataset.plotter.Plotter and :class:~sclab.dataset.processor.Processor can react automatically.

Parameters:

Name	Type	Description	Default
`adata`	`AnnData`	Annotated data matrix to wrap.	required
`name`	`str`	Display name used in exported table filenames. Default is `"SCLabDataset"`.	`'SCLabDataset'`
`counts_layer`	`str`	Name of the raw counts layer. If the layer does not exist it is created from `adata.X`. Default is `"counts"`.	`'counts'`
`copy`	`bool`	If True, work on a copy of `adata`. Default is True.	`True`
`broker`	`EventBroker or None`	Shared event broker. A new one is created if None. Default is None.	`None`

Attributes:

Name	Type	Description
`adata`	`AnnData`	The underlying annotated data matrix.
`obs_table`	`GridBox`	Interactive ipywidget displaying `adata.obs`.
`var_table`	`GridBox`	Interactive ipywidget displaying `adata.var`.
`selected_rows`	`Index`	Index of currently selected observations.

Processor¶

Processor ¶

Processor(dataset: SCLabDataset, plotter: Plotter, *, batch_key: str | None = None)

Bases: EventClient

add_steps ¶

add_steps(_steps: _ProcessorStep | type | list[_ProcessorStep | type] | dict[str, _ProcessorStep | type] | dict[str, list[_ProcessorStep | type]], step_group_name: str = 'Processing')

Add one or more steps to the dataset processor. The steps can be given as a single step instance, a type of step to be instantiated, a list of steps, or a dictionary of step group names to lists of steps.

Parameters:

Name	Type	Description	Default
`_steps`	`_ProcessorStep \| type \| list[_ProcessorStep \| type] \| dict[str, _ProcessorStep \| type] \| dict[str, list[_ProcessorStep \| type]]`	The steps to add to the dataset processor.	required
`step_group_name`	`str`	The name of the step group to add the steps to. If the step group does not exist, it will be created. Defaults to "Processing".	`'Processing'`

Raises:

Type	Description
`ValueError`	If the step has already been added to the dataset processor.

Plotter¶

Plotter ¶

Plotter(dataset: SCLabDataset | None = None, broker: EventBroker | None = None)

Bases: GridBox, EventClient