AttributeError: 'GeoAxesSubplot' object has no attribute '_hold'

[jinlx]

Description

Hi there,

I am trying to do a 2D plot. I get a xarray data 'co' in lat and lon dimension. I want to use co.plot(ax=ax,cmap=WhGrYlRd) to plot it.
However, I got a traceback like this

Traceback (most recent call last):
  File "panel_na.py", line 44, in <module>
    co.plot()
  File "/glade/u/home/lixujin/anaconda3/envs/geoschem/lib/python3.6/site-packages/xarray/plot/plot.py", line 552, in __call__
    return plot(self._da, **kwargs)
  File "/glade/u/home/lixujin/anaconda3/envs/geoschem/lib/python3.6/site-packages/xarray/plot/plot.py", line 187, in plot
    return plotfunc(darray, **kwargs)
  File "/glade/u/home/lixujin/anaconda3/envs/geoschem/lib/python3.6/site-packages/xarray/plot/plot.py", line 854, in newplotfunc
    **kwargs)
  File "/glade/u/home/lixujin/anaconda3/envs/geoschem/lib/python3.6/site-packages/xarray/plot/plot.py", line 1108, in pcolormesh
    primitive = ax.pcolormesh(x, y, z, **kwargs)
  File "/glade/u/home/lixujin/anaconda3/envs/geoschem/lib/python3.6/site-packages/cartopy/mpl/geoaxes.py", line 1451, in pcolormesh
    result = self._pcolormesh_patched(*args, **kwargs)
  File "/glade/u/home/lixujin/anaconda3/envs/geoschem/lib/python3.6/site-packages/cartopy/mpl/geoaxes.py", line 1469, in _pcolormesh_patched
    if not self._hold:
AttributeError: 'GeoAxesSubplot' object has no attribute '_hold

Code to reproduce

# Read data and print out to see the structure
# Pay attention to unit conversion
# Author: Lixu Jin
# Version: GEOS-CHEM 12.2.1, 09/15/2019

# import packages
#from mpl_toolkits.basemap import Basemap

from gamap_colormap import WhGrYlRd
import cartopy.crs as ccrs
import xbpch
import matplotlib.pyplot as plt
import xarray as xr    
import numpy as np
from matplotlib.colors import LogNorm

plt.style.use(['seaborn-talk', 'seaborn-ticks'])

#=============================================================
# Read single dataset
#=============================================================
# Read data
filedir = '/glade/u/home/lixujin/project/GEOS-CHEM/ExtData/HEMCO/GFAS/v2018-09/2018/'
data1 = 'GFAS_201807.nc'
data2 = 'GFAS_201808.nc'
data3 = 'GFAS_201809.nc'
ds1 = xr.open_dataset(filedir + data1)
ds2 = xr.open_dataset(filedir + data2)
ds3 = xr.open_dataset(filedir + data3)
#============================================================
# calculation part
#============================================================
# sum up druing we-can period
co = 60*60*24*1e3*(ds1['cofire'].sum('time') + ds2['cofire'].sum('time') + ds3['cofire'].sum('time'))
co.attrs['unit'] = 'ppm'
print(co)
print(type(co))

# fig setting
# empty map.
fig = plt.figure(figsize=(8,4))
# without this, it would be super wired.
ax = plt.axes(projection=ccrs.PlateCarree())
ax.coastlines()
ax.set_title('global CO during wecan period')

co.plot(ax=ax,cmap=WhGrYlRd)

#fig.plot(cmap='gist_stern') # The ax keyword is particularly useful for subplots. Can be omitted here.
fig.savefig('co_global1.jpg')

[greglucas]

I think your title is misleading, I don't see that error in your output at all?

It doesn't look like you are doing a surface plot to me. I don't know what your dataset is, but here is a link to xarray map examples. Try following that with the contour call.
http://xarray.pydata.org/en/stable/plotting.html#maps

[jinlx]

I think your title is misleading, I don't see that error in your output at all?

It doesn't look like you are doing a surface plot to me. I don't know what your dataset is, but here is a link to xarray map examples. Try following that with the contour call.
http://xarray.pydata.org/en/stable/plotting.html#maps

Sorry, I use the wrong traceback. I have updated it now. Thanks for pointing out it.

[greglucas]

Ah, I was answering the wrong question then :) Now with the new traceback, it looks like this is the same issue as seen here:
#1120

[dopplershift]

I agree that this looks like #1120. @jinlx what versions of cartopy and matplotlib do you have installed? Can you upgrade to more recent versions?

[jinlx]

Here are the versions of my cartopy and matplotlib.
Cartopy 0.16.0
matplotlib 3.1.1
I think you are right. It is caused by old version I use.

[dopplershift]

Well #1120 was fixed in CartoPy 0.17.0, so I'm pretty sure upgrading would solve your problem.

[jinlx]

Well #1120 was fixed in CartoPy 0.17.0, so I'm pretty sure upgrading would solve your problem.

Ryan--Thanks!