matplotlib和pandas在新行上绘制子图

Question

I can plot my pandas dataframe column values as subplots easily in a 1XN matrix.

However, when i want to plot it on a MXN matrix, i get errors.

Example:

df_play = pd.DataFrame({'a':['cat','dog','cat'],
                        'b':['apple','orange','orange'],
                        'c':['boy','boy','girl'],
                        'd':['chair','table','desk']
                       },dtype='category')


fig, axs = plt.subplots(1,len(df_play.columns),figsize=(14,6))
for i,x in enumerate(df_play.columns):
    df_play[x].value_counts().plot(kind='bar',ax=axs[i])

Doing this gives me errors( example i want to view my subplots as a 2X2 matrix):

fig, axs = plt.subplots(2,len(df_play.columns)/2,figsize=(14,6))
for i,x in enumerate(df_play.columns):
    df_play[x].value_counts().plot(kind='bar',ax=axs[i])

AttributeError: 'numpy.ndarray' object has no attribute 'get_figure'

Answer 1

When the subplots are 2-dimensional, subplots returns a figure and a 2-dimensional NumPy array of axes. Therefore, use

axs = axs.ravel()

to make the 2D array of axes 1-dimensional. Then you can use axs[i] to index axs as desired:

for i,x in enumerate(df_play.columns):
    df_play[x].value_counts().plot(kind='bar',ax=axs[i])

axs.ravel() enumerates the axes going from left to right across the rows from the top row down to the bottom row. To enumerate the axes going from top to bottom down the columns from the left-most to the right-most column, use axs.ravel(order='F') .