我有如下图，使用对称对数刻度：

from matplotlib import pyplot as plt
from matplotlib import ticker
(_, ax) = plt.subplots()
ax.plot([-100, -1, -0.5, 0, 0.5, 1, 100], marker='o')
ax.set_yscale('symlog', linthresh=1)
ax.yaxis.set_major_formatter(ticker.LogFormatter(linthresh=1))

我强制让y 轴标签使用标准记法（非科学记数法），但请注意负y 轴上的标签没有负号。该如何修正？

解决方案

Your use of LogFormatter on the y axis actually seems to cause the problem in my opinion: Looking at the source code of it (in particular, line 1013), the LogFormatter simply takes the absolute value of each provided value and then never recovers the sign (it calls self.fix_minus(s) on return, but this is just a correction for "typographically correct" minus signs – which does not have any effect here, since the minus signs have been suppressed before; see above).

Simply using ticker.ScalarFormatter() instead of ticker.LogFormatter(linthresh=1) does seem to fix it for me, however:

完整代码：

from matplotlib import pyplot as plt
from matplotlib import ticker

_, ax = plt.subplots()
ax.plot([-100, -1, -0.5, 0, 0.5, 1, 100], marker="o")
ax.set_yscale("symlog", linthresh=1)
ax.yaxis.set_major_formatter(ticker.ScalarFormatter())