Chord diagrams

using bokeh
visualization
Published

January 17, 2023

Need

Generate quickly such visualizations

image.png

For a larger gallery, a great site Python Graph Gallery

image.png

Using excel

There are examples from https://sites.google.com/site/e90e50fx/home/talent-traffic-chart-with-chord-diagram-in-excel

that we can download and use.

Seems limited to a fixed number of category

Using bokeh - holoviews

https://holoviews.org/reference/elements/bokeh/Chord.html

import pandas as pd
import holoviews as hv
from holoviews import opts, dim
from bokeh.sampledata.les_mis import data

hv.extension('bokeh')
hv.output(size=200)
links = pd.DataFrame(data['links'])
print(links.head(3))
   source  target  value
0       1       0      1
1       2       0      8
2       3       0     10
pd.DataFrame(data['nodes'])
name group
0 Myriel 1
1 Napoleon 1
2 Mlle.Baptistine 1
3 Mme.Magloire 1
4 CountessdeLo 1
... ... ...
72 Toussaint 5
73 Child1 10
74 Child2 10
75 Brujon 4
76 Mme.Hucheloup 8

77 rows × 2 columns

pd.DataFrame(data['links'])
source target value
0 1 0 1
1 2 0 8
2 3 0 10
3 3 2 6
4 4 0 1
... ... ... ...
249 76 66 1
250 76 63 1
251 76 62 1
252 76 48 1
253 76 58 1

254 rows × 3 columns

# libraries
import pandas as pd
import holoviews as hv
from holoviews import opts, dim
from bokeh.sampledata.les_mis import data

# data set
nodes = hv.Dataset(pd.DataFrame(data['nodes']), 'index')
links = pd.DataFrame(data['links'])


# chord diagram
chord = hv.Chord((links, nodes)).select(value=(5, None))
chord.opts(
    opts.Chord(cmap='Category20', edge_cmap='Category20', edge_color=dim('source').str(), 
               labels='name', node_color=dim('index').str()))
links
source target value
0 1 0 1
1 2 0 8
2 3 0 10
3 3 2 6
4 4 0 1
... ... ... ...
249 76 66 1
250 76 63 1
251 76 62 1
252 76 48 1
253 76 58 1

254 rows × 3 columns

with my own data

nodes_df = pd.read_excel('~/vbox/janus/temp/chord diagram.xlsx', sheet_name='nodes')
human_links_df = pd.read_excel('~/vbox/janus/temp/chord diagram.xlsx', sheet_name='links', header=[1], index_col=1 )
human_links_df.drop(columns=human_links_df.columns[0], axis=1,  inplace=True)
nodes_df.head()
name group
0 guillaume 1
1 emilie 1
2 sylvain 1
3 teddy 1
4 gilles 1 
human_links_df.head()
guillaume emilie sylvain teddy gilles gaurav jerome jupyter vscode iolab wsl windows azure ml dataiku gitlab artifactory stormshield disk nas d1 azure blob storage
name
guillaume NaN NaN NaN NaN NaN NaN NaN 9.0 1.0 3.0 7.0 0.0 1.0 1.0 8.0 3.0 5.0 5.0 0.0
emilie NaN NaN NaN NaN NaN NaN NaN 5.0 5.0 5.0 NaN 5.0 NaN 0.0 5.0 NaN 5.0 5.0 NaN
sylvain NaN NaN NaN NaN NaN NaN NaN 2.0 8.0 2.0 8.0 NaN NaN NaN 8.0 3.0 8.0 NaN 2.0
teddy NaN NaN NaN NaN NaN NaN NaN 5.0 5.0 10.0 NaN NaN NaN NaN 5.0 NaN NaN 10.0 NaN
gilles NaN NaN NaN NaN NaN NaN NaN 10.0 NaN 10.0 NaN NaN NaN NaN 5.0 NaN NaN 10.0 NaN 
# we have to transpose human_link to something simpler
liste_nonnan = list(human_links_df[human_links_df.notnull()].stack().index)
liste_refs = []
for (name1, name2) in liste_nonnan:
    # print(f'{name1}, {name2}, {human_links_df.loc[name1,name2]}')
    name1_id = nodes_df.index[nodes_df['name'] == name1].tolist()[0]
    name2_id = nodes_df.index[nodes_df['name'] == name2].tolist()[0]
    value = human_links_df.loc[name1,name2]
    # print(f'{name1_id}, {name2_id}, {human_links_df.loc[name1,name2]}\n')
    if value >0:
        liste_refs.append([name1_id, name2_id, value])
links_df = pd.DataFrame(liste_refs, columns=['source', 'target', 'value'])
links_df.head()
source target value
0 0 7 9.0
1 0 8 1.0
2 0 9 3.0
3 0 10 7.0
4 0 12 1.0 
# data set
nodes = hv.Dataset(nodes_df, 'index')
links = links_df.copy()


# chord diagram
# chord = hv.Chord((links, nodes)).select(value=(5, None))
chord = hv.Chord((links, nodes))
chord.opts(
    opts.Chord(cmap='Category20', edge_cmap='Category20', edge_color=dim('source').str(), 
               labels='name', node_color=dim('index').str()))