How do I plot a step function with Bokeh?

Question

In matplotlib to make a step function you write something like this:

import matplotlib.pyplot as plt

x = [1,2,3,4] 
y = [0.002871972681775004, 0.00514787917410944, 
     0.00863476098280219, 0.012003316194034325]

plt.step(x, y)
plt.show()

How do I make a similar graph with Bokeh?

Answer 1

Bokeh has a Step glyph built-in as of version 0.12.11:

from bokeh.plotting import figure, output_file, show

output_file("line.html")

p = figure(plot_width=400, plot_height=400)

# add a steps renderer
p.step([1, 2, 3, 4, 5], [6, 7, 2, 4, 5], line_width=2, mode="center")

show(p)

Answer 2

You can do it with line and some manual work over the arrays. Take in mind an interleaving of lists for the y axis. Also, we want (I guess) to slightly modify the x-values to accomplish right & left limits. Let me offer the following:

from bokeh.plotting import figure
import numpy as np
f = figure()
#
# Example vectors (given):
x = [1,2,3,4]
y = [0.002871972681775004, 0.00514787917410944, 
     0.00863476098280219, 0.012003316194034325]
#
_midVals = np.diff(x)/2.0
xs = set(x[:-1]-_midVals).union(x[1:]+_midVals)
xl = list(xs)
xl.sort()
xx = xl[1:]+xl[:-1]
xx.sort()
yy = y+y
yy[::2] = y
yy[1::2] = y
#
# assert/print coordinates
assert len(xx)==len(yy)
for _p in zip(xx,yy):
    print _p
#
# Plot!
f.line(x=xx, y=yy)
show(f)

The exit of our "sanity check" print there should be:

(0.5, 0.002871972681775004)
(1.5, 0.002871972681775004)
(1.5, 0.00514787917410944)
(2.5, 0.00514787917410944)
(2.5, 0.00863476098280219)
(3.5, 0.00863476098280219)
(3.5, 0.012003316194034325)
(4.5, 0.012003316194034325)

And the plot:

Hope that helps whoever arrives here.

Answer 3

I've written a little function that does this using line, a little more in the matplotlib style than @Brandt's answer. In a jupyter notebook, this is

def step(fig, xData, yData, color=None, legend=None):
    xx = np.sort(list(xData) + list(xData))
    xx = xx[:-1]
    yy = list(yData) + list(yData)
    yy[::2] = yData
    yy[1::2] = yData
    yy = yy[1:]
    fig.line(xx, yy, color=color, legend=legend)
    return fig
# example
import bokeh.plotting as bok
import bokeh
bokeh.io.output_notebook()

x = [1,2,3,4]
y = [0.002871972681775004, 0.00514787917410944, 
     0.00863476098280219, 0.012003316194034325]
f = bok.figure()
f = step(f, x, y)
bok.show(f)

gives you

Of course you can replace output_notebook() with whatever output formatting you wish.