CODEWITHSUNDEEP
pythondataframegraphicsnetworkxbokeh
Elle Oliver
Elle Oliver

Reputation: 41

Nodes with different colors using bokeh

I recently started using Bokeh for interactive network visualization. I'm plotting coordinates for 50 points, nodes that represent machines. Below is image of how my data is represented and my code. (I only put 14 machines to be simpler).

I've managed to plot the points but I have a question that I didn't find anywhere a specific solution. For some machines I have its temperature, but others no. How can I make the machines that I have the temperature information have a different color?

Like, the ones that I have this information be red, and the others that I don't have the information be blue? All tutorials that I found about changing the nodes color involves palettes, but this wouldn't have much use for me now.

    import pandas
    from bokeh.io import output_notebook, show, save
    from bokeh.io import output_notebook, show, save
    from bokeh.models import Range1d, Circle, ColumnDataSource, MultiLine
    from bokeh.plotting import figure
    output_notebook()

    from bokeh.plotting import ColumnDataSource, figure, output_file, show

    df = pandas.read_excel('Pasta1.xlsx', engine='openpyxl')

    source = ColumnDataSource(data=dict(x = df['x'] ,y = df['y']))

    TOOLTIPS = [("index", "$index"),("(x,y)", "($x, $y)")]

    p = figure(width=1000, height=500, tooltips=TOOLTIPS,title="Redes")

    p.circle('x', 'y', size=10,fill_color='red', source=source)

    show(p)

enter image description here enter image description here

Upvotes: 0

Views: 1003

Answers (1)

mosc9575
mosc9575

Reputation: 6337

The solutions is to use the color keyword of the p.circle() and pass a list (or array) instead of a static value. If you now the rules for your color it should be easy to pass the information to your plot.

The keyword fill_color also accepts lists (or arrays), if you prefere this.

Example

The example below creates the color column in the pandas DataFrame first, using np.where(). This can be done also by hand or using other technics.

import numpy as np
import pandas as pd

from bokeh.io import output_notebook, show, save
from bokeh.models import Range1d, Circle, ColumnDataSource, MultiLine
from bokeh.plotting import figure
output_notebook()


df = pd.DataFrame({
    'machine':['J'+str(i) for i in range(13)], 
    'x':list(range(13)),
    'y':list(range(13)),
    'Temp' : [np.nan, 32, np.nan, 33, np.nan, np.nan, np.nan, np.nan, 35, np.nan, np.nan, 32, np.nan]
})
df['color'] = np.where(df['Temp'].isna(), 'blue', 'red')
source = ColumnDataSource(df)

TOOLTIPS = [("index", "$index"),("(x,y)", "($x, $y)"), ('name', "@machine")]

p = figure(width=500, height=500, tooltips=TOOLTIPS,title="Redes")
p.circle(x='x', y='y', size=10, color='color', source=source)
show(p)

Output

colored circles

Upvotes: 1

Related Questions