Figure factory: Plot a scalar field with geographical coordinates on map with projection

I have an array of global temperatures and associated longitude and latitude coordinates arrays. I would like to plot the temperatures as a heatmap on a map with custom projection and showing coastlines.
A working code example using matplotlib is:

import matplotlib.pyplot as plt
import cartopy.crs as ccrs

f, ax = plt.subplots(subplot_kw={'projection': ccrs.Robinson()})
p = ax.pcolormesh(longitude, latitude, temperature, cmap='RdBu_r', transform=ccrs.PlateCarree())
ax.coastlines()
f.colorbar(p, orientation='horizontal')

The output figure from code block above is (data available here):

At the moment I was only able to use plotly for a simple heatmap with the following code:

from plotly.offline import iplot
import plotly.graph_objs as go

layout = go.Layout(geo={'projection': {'type': 'robinson'}})
data = go.Heatmap(x=longitude, y=latitude, z=temperature)

fig = go.Figure(data=[data], layout=layout)
iplot(fig)

As next figure shows, main issues are:

• the projection is not working;

• no coastlines are shown.
  

提供された plotly.graph_objs の Heatmap 例から始め、matplotlib/cartopy の動作と比較してください。経度/緯度の温度データが、要求された投影法と表示可能な海岸線を備えた heatmap としてレンダリングされれば、作業は完了です。