How do I save matplotlib backend opts without calling hv.extension('matplotlib')?

hvPlot
1 
import hvplot.xarray
import holoviews as hv
import xarray as xr

ds = xr.tutorial.open_dataset('air_temperature').isel(time=slice(0, 3))
plot = ds.hvplot('lon', 'lat', dynamic=False).opts(fig_size=300)
hv.save(plot, 'testmaapl.html', resources='inline', backend='matplotlib')

I get these warnings:

WARNING:param.main: Option 'fig_size' for Image type not valid for selected backend ('bokeh'). Option only applies to following backends: ['matplotlib']
WARNING:param.main: Option 'fig_size' for Image type not valid for selected backend ('bokeh'). Option only applies to following backends: ['matplotlib']
WARNING:param.main: Option 'fig_size' for Image type not valid for selected backend ('bokeh'). Option only applies to following backends: ['matplotlib']

Because I flip flop between bokeh and matplotlib backends for different plots

e.g.
hv.extension(‘matplotlib’)
…plot some map with matplotlib

hv.extension(‘bokeh’)
…plot some map with bokeh

hv.extension(‘matplotlib’)
…again plot some map with matplotlib

2

You can provide the backend to the opts call:

plot = ds.hvplot('lon', 'lat', dynamic=False).opts(fig_size=300, backend='matplotlib')

You may also have to import the matplotlib backend though:

import holoviews.plotting.mpl
1 Like