Can't update multi_line datasource dynamically

I have the following code to update a multi-line plot dynamically:

xs, ys = get_xs_ys_for(event)
source = ColumnDataSource(dict(xs=xs, ys=ys))

def ticker_change(attrname, old, new):
    xs, ys = get_xs_ys_for(new)
    source.data = dict(x=xs, ys=ys)
ticker.on_change('value', ticker_change)

fig = figure(title='Individual Lines', height=450)
plot = fig.multi_line(xs, ys, source=source, line_width=5, line_color=Spectral11)

However, the update is not reflected in the multi-line plot when I change the ticker (it keeps the old lines). I’m wondering if it is a bug (because it works with one line (using fig.line())?

[bryevdv]

You can set the line color but you need to put the colors in the source. Behaviours like this are why this mixed list literal / explicit source usage is being disallowed. In brief the problem is this: the original glyph method call constructs a column for line_color but your update function only has x and y columns. So then the plot looks for a (now non-existent) column for line color, and breaks.

[birdsarah]

Right. OK that was the point I was trying to make I just missed the line_color thing being wrong too.

So this is wrong:

plot = fig.multi_line(xs, ys, source=source, line_width=5, line_color=Spectral11)

this is wrong:

plot = fig.multi_line(xs='xs', ys='ys', source=source, line_width=5, line_color=Spectral11)

this is right:

plot = fig.multi_line(xs='xs', ys='ys', source=source, line_width=5, line_color='color')

ALTERNATIVELY

You could do this which pulls the source that’s been created for you out:

xs, ys = get_xs_ys_for(all_events[0])
fig = figure(title='Individual CPUs', height=450)
lines = fig.multi_line(xs=xs, ys=ys, line_width=5, line_color=Spectral11)
source = lines.data_source

def ticker_change_bad(attrname, old, new):
   # DON'T DO THIS
    xs, ys = get_xs_ys_for(new)
    source.data['xs'] = xs
    source.data['ys'] = ys

def ticker_change_good(attrname, old, new):
    xs, ys = get_xs_ys_for(new)
    source.data = dict(xs=xs, ys=ys, line_color=Spectral11)

# The two ticker change methods should be equivalent. I'm not 100% sure what the name of the #column will be when the source is auto-generated for you.

ticker.on_change('value', ticker_change_good)