In matplotlib, Is there a way to set gridlines below bars/lines/patches while retaining ticklabels above?

Question

Related to Matplotlib: draw grid lines behind other graph elements, but nothing there worked for me.

I have the following plot where I want to hide the gridlines under the red line while retaining the labels on top of the red line:

import numpy as np
import matplotlib.pyplot as plt

#plot
r = np.arange(0, 3.0, 0.01)
theta = 2 * np.pi * r
ax = plt.subplot(111, polar=True)
ax.plot(theta, r, color='r', linewidth=20)
ax.set_rmax(2.0)
ax.grid(True, lw=2)
#set labels
label_pos = np.linspace(0.0, 2 * np.pi, 6, endpoint=False)
ax.set_xticks(label_pos)
label_cols = ['Label ' + str(num) for num in np.arange(6)]
ax.set_xticklabels(label_cols, size=24)

I can get the red line on top with ax.set_axisbelow(True).

But I can't find a way to keep the red line on top of the gridlines while retaining the labels on top of the red line. Adding zorder=-1 to the plot command, puts the red line in the bottom even if I add ax.set_axisbelow(True). ax.set_zorder(-1)) has not worked so far either.

How can I get the grid lines in the bottom (lowest zorder) followed by the red line and then the labels on top of the red line?

Batter

Answer 1

You can always plot the grid manually:

import numpy as np
import matplotlib.pyplot as plt

#plot
r = np.arange(0, 3.0, 0.01)
theta = 2 * np.pi * r
rmax = 2.0
n_th = 6
th_pos = np.linspace(0.0, 2 * np.pi, n_th, endpoint=False)
n_r = 5
r_pos = np.linspace(0, rmax, n_r)


ax = plt.subplot(111, polar=True)

## Plot the grid    
for pos in th_pos:
    ax.plot([th_pos]*2, [0, rmax], 'k:', lw=2)
for pos in r_pos[1:-1]:
    x = np.linspace(0, 2*np.pi, 50)
    y = np.zeros(50)+pos
    ax.plot(x, y, 'k:', lw=2)

## Plot your data
ax.plot(theta, r, color='r', linewidth=20)
ax.set_rmax(rmax)
ax.grid(False)

#set ticks and labels
ax.set_xticks(th_pos)
label_cols = ['Label ' + str(num) for num in np.arange(n_th)]
ax.set_xticklabels(label_cols, size=24)
ax.set_yticks(r_pos[1:])


plt.show()

Draper

Thanks! I was thinking about plotting them individually, but was hoping for a simple zorder rearrangement. This is not too much extra code though, will accept in a couple of days if nothing simpler comes up. – Batter

Sounds good :-) I thought you might prefer a simpler solution. – Draper

@cheflo have you found a better solution? – Draper