Improve the Plotting section of the tutorial

[alixdamman]

Like in Pandas doc ( http://pandas.pydata.org/pandas-docs/stable/visualization.html )

[gdementen]

Given #925, a lot of plots are much easier (or even possible in some cases) to do now. A section of the tutorial showcasing that would be very helpful I think.

[gdementen]

Here are a few examples of plots which look "good enough" using #925.

pop = load_example_data('demography').pop
pop[2016].plot.bar(x='age', y='sex', xticks=[0, 20, 40, 60, 80, 100], stacked=True, subplots=('nat', 'geo'))
pop.sum('age', 'sex').plot(x='time', subplots='nat')
pop[2016].plot(x='FO', y='BE', subplots=('geo', 'sex'))

pop[2016].plot.box(by=('nat', 'sex'))
pop[2016].sum('nat').plot.box()

pop[2016].plot.hist()
pop[2016].plot.hist(by='nat', stacked=True)

pop[2016].plot.kde(by='nat')

pop.plot.scatter(x=1991, y=2016, subplots=('geo', 'nat'), sharex=False)
pop[2016].sum('sex', 'geo').plot.scatter(x='FO', y='BE')
pop[2016].plot.scatter(x='FO', y='BE', subplots=('geo', 'sex'))

pop[2016].plot(x='FO', y='BE', subplots=('geo', 'sex'))

pop[2016].sum('age', 'sex').plot.pie(subplots='geo')

pop.sum('age').plot.area(x='time')
pop.sum('age').sum('sex').plot.area(x='time', subplots='geo')
pop.sum('age').sum('sex').plot.area(x='time', subplots='nat')