[FIXED] Tick labels position

Issue

I want to plot a roc curve, but the tick label 0.0 appears at both axes, I removed one tick label by directly setting the labels:

pl.gca().set_yticks([0.2, 0.4, 0.6, 0.8, 1.0])
pl.gca().set_xticks([0.0, 0.2, 0.4, 0.6, 0.8, 1.0])

How can the tick label "0.0" of the x-axis aligned with the y-axis? The label should be moved to the left border of the y-axis, that it begins at the same vertical position as the other tick labels in the y-axis.

Solution

I think you want to prune x-axis:

#!/usr/bin/env python3

import matplotlib
from matplotlib import pyplot as plt
from matplotlib.ticker import MaxNLocator

data = range(5)


fig = plt.figure()
ax = fig.add_subplot(111)

ax.plot(data,data)

ax.xaxis.set_major_locator(MaxNLocator(5, prune='lower'))
ax.yaxis.set_major_locator(MaxNLocator(4))

fig.savefig("1.png")

Edit:

Sad but true: matplotlib was not meant for a crossed axes 2D plots. If You sure the zero for both axis is at the lower left box corner than I suggest to put it there manually:

#!/usr/bin/env python3

import matplotlib
from matplotlib import pyplot as plt
from matplotlib.ticker import MaxNLocator

data = range(5)


fig = plt.figure()
ax = fig.add_subplot(111)

ax.plot(data,data)

ax.xaxis.set_major_locator(MaxNLocator(5, prune='lower'))
ax.yaxis.set_major_locator(MaxNLocator(4, prune='lower'))

fig.tight_layout()

ax.text(-0.01, -0.02,
        "0",
        horizontalalignment = 'center',
        verticalalignment = 'center',
        transform = ax.transAxes)

fig.savefig("1.png")

Here one can adjust the zero position manually.

Personally I prune x or y axis depending on situation, and happy with that.

Answered By - Adobe