hv_anndata.ManifoldMap

class hv_anndata.ManifoldMap(**params)

Interactive manifold map application for exploring AnnData objects.

This application provides widgets to select dimensionality reduction methods, dimensions for x and y axes, coloring variables, and display options.

Parameters:
adata

AnnData object to visualize

reduction

Initial dimension reduction method to use

color_by_dim

Color by dimension, one of ‘obs’ (default) or ‘cols.

color_by

Initial variable to use for coloring

colormap

Initial colormap. Auto-updates based on data type. Options: - Categorical: “Glasbey Cat10”, “Cat20”, “Glasbey cool” - Continuous: “Viridis”, “Fire”, “Blues” Custom colormaps are not yet supported.

datashade

Whether to enable datashading

width

Minimum width of the plot. If responsive is True, this is the minimum width.

height

Minimum height of the plot. If responsive is True, this is the minimum height.

show_labels

Whether to show labels

show_widgets

Whether to show control widgets

responsive

Whether to make the plot size-responsive

plot_opts

HoloViews plot options for the manifoldmap plot

labeller_opts

Operation and plot options for the labeller

Parameter Definitions

adata = ClassSelector(allow_None=True, class_=<class 'anndata.AnnData'>, label='Adata')

AnnData object to visualize

reduction = Selector(label='Reduction', names={}, objects=[])

Dimension reduction method

x_axis = Selector(label='X axis', names={}, objects=[])

y_axis = Selector(label='Y axis', names={}, objects=[])

color_by_dim = Selector(default='obs', names={'Observations': 'obs', 'Variables': 'cols'}, objects=['obs', 'cols'])

color_by = Selector(label='Color by', names={}, objects=[])

Coloring variable

colormap = Selector(allow_None=True, label='Colormap', names={}, objects=['Glasbey Cat10', 'Cat20', 'Glasbey cool', 'Viridis', 'Fire', 'Blues'])

Initial colormap name. Auto-selects default if type mismatched

datashade = Boolean(default=True)

Whether to enable datashading

var_reference = String(allow_None=True, label='Var reference')

Column name in .var to use for populating the variable names, default to the index names if not set.

width = Integer(default=300, inclusive_bounds=(True, True), label='Width')

Minimum width of the plot

height = Integer(default=300, inclusive_bounds=(True, True), label='Height')

Minimum height of the plot

show_labels = Boolean(default=False)

Whether to show labels

show_widgets = Boolean(default=True, label='Show widgets')

Whether to show control widgets

ls = ClassSelector(allow_None=True, class_=<class 'holoviews.selection.link_selections'>, label='Ls')

streams = List(bounds=(0, None), default=[], label='Streams')

List of streams to use for dynamic updates

responsive = Boolean(default=True, label='Responsive')

Whether to make the plot size-responsive

plot_opts = Dict(class_=<class 'dict'>, default={}, label='Plot opts')

HoloViews plot options for the manifoldmap plot

legend_position = Selector(default='bottom_right', label='Legend position', names={}, objects=['top_left', 'top_center', 'top_right', 'center_left', 'center', 'center_right', 'bottom_left', 'bottom_center', 'bottom_right', 'right', 'left'])

Bokeh legend position

legend_alpha = Number(bounds=(0, 1), default=0.6, inclusive_bounds=(True, True), label='Legend alpha', step=0.1)

Legend background opacity

legend_font_size = Integer(bounds=(5, 14), default=8, inclusive_bounds=(True, True), label='Legend font size')

Legend label font size in pt

legend_ncols = Integer(bounds=(1, 4), default=1, inclusive_bounds=(True, True), label='Legend ncols')

Number of legend columns

labeller_opts = Dict(class_=<class 'dict'>, default={}, label='Labeller opts')

Operation and plot options for the labeller

_replot = Event(default=False, label=' replot')

_categorical = Boolean(default=False, label=' categorical')

Attributes

Methods

create_plot(*, dr_key, x_value, y_value, ...)

Create a manifold map plot with the specified parameters.

get_dim_labels(dr_key)

Get labels for each dimension in a reduction method.

get_reduction_label(dr_key)

Get a display label for a dimension reduction key.