Instead of drawing the subcoordinate grouping from the Element.group you can now specify a multi-dimensional NdOverlay and draw the group from one of the dimensions, e.g.:
import numpy as np
import holoviews as hv
hv.extension('bokeh')
data = {
('EEG', f'EEG {i}'): hv.Curve(
np.random.randn(100).cumsum()+i
).opts(subcoordinate_y=True)
for i in range(3)
}
data.update({
('MEG', f'MEG {i}'): hv.Curve(
-np.random.randn(100).cumsum()+i
).opts(subcoordinate_y=True)
for i in range(3)
})
hv.NdOverlay(data, ['Group', 'Channel']).opts(
responsive=True, height=400, subcoordinate_group='Group'
)
Instead of drawing the subcoordinate grouping from the
Element.group
you can now specify a multi-dimensional NdOverlay and draw the group from one of the dimensions, e.g.: