How to use pandas plotting with subplots. ax=df.plot() or df.plot(ax=ax1). whats the difference

Question

I have created two axes in a figure using matplotlib and tried to plot data. One method worked and one did not. My question: why did it not work and what's the difference between the two?

The one which didn't work:

fig,ax=plt.subplots(2,figsize=(15,6))
ax[0]=df['Global_Sales'].head(10).plot(kind='bar')
ax[0].set_xlabel('different games')
ax[0].set_ylabel('Sales')
ax[1]=df['Critic_Score'].head(10).plot(kind='bar')
ax[1].set_xlabel('different games')
ax[1].set_ylabel('Critic Score')
plt.tight_layout()
plt.show()

Here the first axes remains blank and the second axes gets overwritten

The one which did work:

fig,ax=plt.subplots(2,figsize=(15,6))
df['Global_Sales'].head(10).plot(kind='bar',ax=ax[0])
ax[0].set_xlabel('different games')
ax[0].set_ylabel('Sales')
df['Critic_Score'].head(10).plot(kind='bar',ax=ax[1])
ax[1].set_xlabel('different games')
ax[1].set_ylabel('Critic Score')
plt.tight_layout()
plt.show()

showed both the graphs on both axes properly.

Answer 1

Your first example did not produce the output you were expecting because you did not pass the axes objects that you generated using the plt.subplots call to the pandas .plot call. Your second example produced the output you were expecting becuase you did pass the axes objects that you generated using the plt.subplots call to the pandas .plot call.

In short, if you want a pandas .plot call to plot on a particular axis (say ax1 ), you need to pass that axis object into the .plot call as a keyword argument: df.plot(ax=ax1)