Draw horizontal lines from x=0 to data points in matplotlib scatterplot (horizontal stem plot)

问题

Consider the follwing plot:

produced by this function:

def timeDiffPlot(dataA, dataB, saveto=None, leg=None):
    labels = list(dataA["graph"])
    figure(figsize=screenMedium)
    ax = gca()
    ax.grid(True)
    xi = range(len(labels))
    rtsA = dataA["running"] / 1000.0 # running time in seconds
    rtsB = dataB["running"] / 1000.0 # running time in seconds
    rtsDiff = rtsB - rtsA
    ax.scatter(rtsDiff, xi, color='r', marker='^')
    ax.scatter
    ax.set_yticks(range(len(labels)))
    ax.set_yticklabels(labels)
    ax.set_xscale('log')
    plt.xlim(timeLimits)
    if leg:
        legend(leg)
    plt.draw()
    if saveto:
        plt.savefig(saveto, transparent=True, bbox_inches="tight")

What matters here is the positive or negative difference of the values to x = 0. It would be nice to visualize this more clearly, e.g.

• emphasize the x=0 axis
• draw a line from x=0 to the plot marker

Can this be done with matplotlib? What code would need to be added?

回答1:

As pointed out by Rutger Kassies, there are actually some "stem" functions that automate the "manual" method from my other answer. The function for horizontal stem lines is hlines() (vlines() is for vertical stem bars):

import numpy
from matplotlib import pyplot

x_arr = numpy.random.random(10)-0.5; y_arr = numpy.arange(10)

pyplot.hlines(y_arr, 0, x_arr, color='red')  # Stems
pyplot.plot(x_arr, y_arr, 'D')  # Stem ends
pyplot.plot([0, 0], [y_arr.min(), y_arr.max()], '--')  # Middle bar

The documentation for hlines() is on the Matplotlib website.

回答2:

(See my other answer, for a faster solution.)

Matplotlib offers vertical "stem" bars: http://matplotlib.org/api/pyplot_api.html#matplotlib.pyplot.stem. However, I can't find a horizontal equivalent of stem().

One can nonetheless draw horizontal stem bars quite easily, through repeated plot() calls (one for each stem). Here is an example

import numpy
from matplotlib.pyplot import plot

x_arr = numpy.random.random(10)-0.5; y_arr = numpy.arange(10)

# Stems:
for (x, y) in zip(x_arr, y_arr):
    plot([0, x], [y, y], color='red')
# Stem ends:
plot(x_arr, y_arr, 'D')
# Middle bar:
plot([0, 0], [y_arr.min(), y_arr.max()], '--')

with the following result:

Note, however, that drawing bars from x = 0 does not make sense when x is on a log scale, as David Zwicker pointed out, since x = 0 is infinitely far on the left of the x axis.

来源：https://stackoverflow.com/questions/14957567/draw-horizontal-lines-from-x-0-to-data-points-in-matplotlib-scatterplot-horizon