Wrong Tracker values on a 3D histogram

Question

Here is some code which displays a 3D histogram. However, the tracker in the lower-right corner does not correctly display the location of the mouse.

The tracker says x = e while the mouse is clearly over c. And the tracker says z = 01-02. What's up with that? (The z tracker value seems to be controlled by the y-axis formatter.)

How can the code be fixed?

import matplotlib.pyplot as plt
import numpy as np
import mpl_toolkits.mplot3d.axes3d as axes3d
import matplotlib.dates as mdates
import matplotlib.ticker as ticker
import datetime as dt
import random

np.random.seed(0)
fig = plt.figure()
ax = fig.add_subplot(1, 1, 1, projection = '3d')

cmap = plt.get_cmap('RdBu')
event_labels = 'abcdefghij'
events = range(len(event_labels))
label_map = dict(zip(events,event_labels))

dates = mdates.drange(dt.datetime(2012, 10, 1),
                      dt.datetime(2012, 10, 10),
                      dt.timedelta(days = 1))
events_list = [(random.choice(dates), random.choice(events))
               for i in range(50)]

event_array, date_array = zip(*events_list)

# Much of the code below comes from 
# http://matplotlib.org/examples/mplot3d/hist3d_demo.html
hist, xedges, yedges = np.histogram2d(date_array, event_array)

elements = (len(xedges)-1) * (len(yedges)-1)
xpos, ypos = np.meshgrid(xedges[:-1]+0.25, yedges[:-1]+0.25)

xpos = xpos.flatten()
ypos = ypos.flatten()
zpos = np.zeros(elements)
dx = 0.5 * np.ones_like(zpos)
dy = dx.copy()
dz = hist.flatten()

ax.bar3d(xpos, ypos, zpos, dx, dy, dz, color='b')

xfmt = ticker.FuncFormatter(lambda x, pos: label_map[int(x)])
yfmt = mdates.DateFormatter('%m-%d')
zfmt = ticker.FuncFormatter(lambda z, pos: str(z))
ax.w_xaxis.set_major_formatter(xfmt)
ax.w_yaxis.set_major_formatter(yfmt)
ax.w_zaxis.set_major_formatter(zfmt)

ax.fmt_xdata = xfmt
ax.fmt_ydata = yfmt
ax.fmt_zdata = zfmt
plt.show()

Answer 1

First, you might change your xaxis formatter, currently it is throwing a lot of exceptions out to the interpreter when the mouse goes to the edges of the chart:

You might change this:

 xfmt = ticker.FuncFormatter(lambda x, pos: label_map[int(x)])

To something like this:

def xformatter(x,pos):
   try:
       val = label_map[int(x)]
   except:
       val = "None"
   return val
...
xfmt = ticker.FuncFormatter(xformatter)

Next, I found there is an error in the actual mpl_toolkits.mplot3d

If you take a look at line 320 in the \Lib\site-packages\mpl_toolkits\mplot3d\axis3d.py

You'll find this typo:

class YAxis(Axis):
    def get_data_interval(self):
        'return the Interval instance for this axis data limits'
        return self.axes.xy_dataLim.intervaly

class ZAxis(Axis):
    def get_data_interval(self):
        'return the Interval instance for this axis data limits'
        return self.axes.zz_dataLim.intervalx

Note how self.axes.zz_dataLim.intervalx should be: self.axes.zz_dataLim.intervalz

Probably, you need to report these problems to the developer.