交互式 window 中鼠标 position 的不同格式化程序

Question

I have a plot where in abscissa I have the year 2022, and in ordinate a function of time

In [215]: import matplotlib.pyplot as plt
     ...: from matplotlib.dates import MonthLocator, DateFormatter
     ...: from numpy import linspace, sin
     ...: 
     ...: T0 = (2022-1970)*365.25 ; T1 = T0+365 ; T = np.linspace(T0, T1, 366)
     ...: w = 6.28/40
     ...: 
     ...: fig, ax = plt.subplots(figsize=(14,5), constrained_layout=1)
     ...: ax.plot(T,40+15*np.sin(w*T-0.2))
     ...: ax.xaxis.set_major_locator(MonthLocator(bymonth=range(13)))
     ...: ax.xaxis.set_major_formatter(DateFormatter("%b %Y"))
     ...: 
     ...: plt.show()

As you can see from the code, the function is plotted with a one day resolution, but because of my choice for the formatting of x-tick-labels when I move the mouse over the interactive window what I see in the upper right corner is just, eg, "x=May 2022 y=43.31".

Is it possible to have a different formatter for the interactive window display of the mouse position?

Answer 1

ax.fmt_xdata controls the formatting of the coordinate in the status bar. By default, it uses the same format as the x-axis, but you can assign a more specific one.

Here is an example:

import matplotlib.pyplot as plt
from matplotlib.dates import MonthLocator, DateFormatter
import numpy as np

T0 = (2022 - 1970) * 365.25
T1 = T0 + 365
T = np.linspace(T0, T1, 366)
w = 6.28 / 40

fig, ax = plt.subplots(figsize=(14, 5), constrained_layout=1)
ax.plot(T, 40 + 15 * np.sin(w * T - 0.2))
ax.xaxis.set_major_locator(MonthLocator(bymonth=range(13)))
ax.xaxis.set_major_formatter(DateFormatter("%b %Y"))
ax.fmt_xdata = DateFormatter("%d %b %Y")
plt.show()

You could even go one step further, and override the ax.format_coord function. This is a function, given an x and y value, to produce the full coordinate string to be displayed in the status bar. By default, it just combines ax.fmt_xdata and ax.fmt_ydata .

Also the mplcursors library might be of interest. With it, you can obtain the xy coordinates of the nearest point on the curve, and show that information as an annotation, or as overriding the status bar.