How to set colors when using subplots

hvPlot
1

I’m plotting some data in a histogram and find that I have too much to see in 1 graph.

When I add the subplots=True parameter, I can see my a plot per category, but now every plot has the same color. Preserving color helps viewers identify similar categories across plots.

Is it possible to use subplots=True and keep the automatic color by category?

Not sure if this is a case of not understanding the documentation, a missing feature, or something else.

Below I do show one example of how to color a plot by creating a unique plot per category in a for loop then using Colorcet to set a color. Seems to me there should be a better way. Is there?

Code example available at hvplot_example.ipynb · GitHub

# load hvplot and default bokeh extension
import hvplot.pandas
# load some sample data
import hvplot.sample_data
# colormaps
import colorcet as cc

df = hvplot.sample_data.airline_flights.read()

# Notice how each day of the week automatically is assigned a unique color?
df.hvplot.hist('arr_time',
               by='dayofweek'
              )

# Now when plotting each day of the week on their own, colors are gone and we're back to the default blue.
#
# Where did my colors go? How to preserve them?
df.hvplot.hist('arr_time',
               by='dayofweek',
               subplots=True,
              )

# It is possible to build each individual day of the week plot and color each one individually.
all_p = None
for dow, c in zip(df.dayofweek.unique(), cc.b_glasbey_bw_minc_20[:len(df.dayofweek.unique())]):
    p = df[df.dayofweek == dow].hvplot.hist('arr_time',
                                            by='dayofweek',
                                            color=c,
                                            title='dayofweek: ' + str(dow)
                                           )
    all_p = p if all_p is None else all_p + p
all_p.cols(2)

Screen Shot 2023-11-10 at 16.20.01
Screen Shot 2023-11-10 at 16.20.012072×2424 453 KB

2

Something like this:

import hvplot.pandas
import hvplot.sample_data
import colorcet as cc

df = hvplot.sample_data.airline_flights.read()

plots = df.hvplot.hist("arr_time", by="dayofweek", subplots=True)

for plot, color in zip(plots, cc.b_glasbey_bw_minc_20):
    plot.opts(color=color)

plots.cols(2)
3

Thanks! Very clean and much better than the method I came up with.

Is something like your example in the documentation? If not, I think providing something like it would be helpful for others.