Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Already on GitHub? Sign in to your account

Problem with subplot / matplotlib.dates interaction #1086

DamienIrving opened this Issue Aug 15, 2012 · 1 comment


None yet
2 participants

I've been plotting timeseries data using the matplotlib.dates module and have come across an issue when using it in conjunction with the subplot command.

For figures with greater than one subplot in a particular column, the time (or x) axis ticks and their labels are only printed on the final subplot that is plotted, and are missing on all other subplots in that column (see example code below). The only exception is when the time axis is identical for each plot in the column (you can test this by editing my example code at ###) - then all time axis ticks and their labels are displayed correctly.

Is this a bug in the subplot source code, or am I missing something?

from pylab import *
from matplotlib.dates import YEARLY, DateFormatter, rrulewrapper, RRuleLocator, drange
import datetime

tick every 5th easter

rule = rrulewrapper(YEARLY, byeaster=1, interval=5)
loc = RRuleLocator(rule)
formatter = DateFormatter('%m/%d/%y')

data for subplot 1

date1 = datetime.date( 1952, 1, 1 )
date2 = datetime.date( 2004, 4, 12 )
delta = datetime.timedelta(days=100)

dates = drange(date1, date2, delta)
s = rand(len(dates)) # make up some random y values

plot subplot 1

ax1 = subplot(211)
plot_date(dates, s)
labels = ax1.get_xticklabels()
setp(labels, rotation=30, fontsize=10)

data for subplot 2

date1 = datetime.date( 2052, 1, 1 ) ###( 1952, 1, 1 )
date2 = datetime.date( 2104, 4, 12 ) ###( 2004, 4, 12 )

dates = drange(date1, date2, delta)
s = rand(len(dates)) # make up some random y values

plot subplot 2

ax2 = subplot(212)
plot_date(dates, s)
labels = ax2.get_xticklabels()
setp(labels, rotation=30, fontsize=10)



efiring commented Aug 15, 2012

You are missing something, but it is something that is quite non-intuitive and easy to miss: Locators can't be shared among axes. The set_major_locator() method assigns its axis to that Locator, overwriting any axis that was previously assigned.

@efiring efiring closed this Aug 15, 2012

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment