pandas dataframe：子图如何调整每个数字的图例？

Question

I have 2 dataframes df1 and df2 with column date as index:

       df1 = {'Confirmed':[3,0,0,9],
              'Date' : ['2020-03-20','2020-03-21','2020-03-22','2020-03-23'],
               'Recovered' : [0,0,1,2]
}
       # ratio Recovered / Confirmed 
      df2  = {  'Ratio': [0.00,2.77,2.43,7.95],
                'Date' : ['2020-03-20','2020-03-21','2020-03-22','2020-03-23']

}

when I plot them separately:

# plot  df1  bar 
ax  = df1.plot(

    kind='bar' , figsize=(20,7), grid=True , legend= False,
    color=('r','g'),fontsize=15,
    secondary_y=False,mark_right=False,
    title = 'Covid-19 Madagascar',
    yticks = range (0,30,2)


)
ax.xaxis.set_major_formatter(plt.FixedFormatter(df1.index.to_series().dt.strftime("%Y-%m-%d")))
ax.set_ylabel('Isan\'ny olona',fontsize=15)
ax.set_yticklabels("" , minor =True)
# ax.yaxis.set_major_locator(plt.MinNLocator(1))
plt.legend(loc='upper center')
plt.show()

the result:

now for df2:

# plot  second  Ratio line 
ax  = df2.plot(

    kind='line' , figsize=(20,7), grid=True , legend= False,
    color=('b'),fontsize=15,
    mark_right=False,
    title = 'Covid-19 Madagascar'
    # yticks = range (0,1)


)
# ax.xaxis.set_major_formatter(plt.FixedFormatter(df2.index.to_series().dt.strftime("%Y-%m-%d")))
ax.set_ylabel('Isan-jato %',fontsize=15)
ax.set_yticklabels("" , minor =True)
# ax.yaxis.set_major_locator(plt.MinNLocator(1))
plt.legend(loc='upper center')
plt.show()

the result is:

as you can see when i plotted separately, the result is ok. But when I would like to add those figures in one by using subplot, I have a problem with the legend position, it hides the title.

fig , axes = plt.subplots(ncols=1,nrows=2,constrained_layout=True)
# -------------------------- plot first bar ------------------------------------------- 
ax  = df1.plot(
    ax = axes[0],
    kind='bar' , figsize=(20,15) , grid=True , legend= False,
    color=('r','g'),fontsize=15,
    secondary_y=False,mark_right=False,
    title = 'Covid-19 Madagascar',
    yticks = range (0,30,2)


)

ax.xaxis.set_major_formatter(plt.FixedFormatter(df1.index.to_series().dt.strftime("%Y-%m-%d")))
ax.set_ylabel('Isan\'ny olona',fontsize=17)
ax.set_yticklabels("" , minor =True)
# plt.legend(loc='upper left')
# handles, labels = ax.get_legend_handles_labels()
fig.legend(loc='upper center',prop={'size':15})


# --------------------- plot  second line -----------------------------
ax = df2.plot(
    ax = axes[1],
    kind='line' , figsize=(20,15) ,grid=True , legend= False,
    color=('b'),fontsize=15,
    mark_right=False,


)
# ax.xaxis.set_major_formatter(plt.FixedFormatter(df2.index.to_series().dt.strftime("%Y-%m-%d")))
ax.set_ylabel('Isan-jato %',fontsize=17)
ax.set_yticklabels("" , minor =True)


plt.legend(loc='upper center',prop={'size':15})

plt.show()

how can we do to duplicate the same figures as separately with subplot

Answer 1

You can move the legend with;

ax.legend(bbox_to_anchor=(1.0, 0.75))

Play with the numbers until it's where you want.

Answer 2

Or you could pass the relative coordinates as the loc parameter, instead of 'upper center' :

fig.legend(loc=(0.47, 0.92), prop={'size':15})