如何使用 Matplotlib 制作逐行派 plot？

Question

My dataframe looks like -

state      value1       value2       value3
  a          20           65           15
  b          35           35           30
  c          30           25           45

I want a pie chart for each state( may be 20 states are present). My code is given below -

For pie chart -

    for ind in df.index:
        fig, ax = plt.subplots(1,1)
        fig.set_size_inches(5,5)
        df.iloc[ind].plot(kind='pie', ax=ax, autopct='%1.1f%%')
        ax.set_ylabel('')
        ax.set_xlabel('')

But it does not work.

Answer 1

Probably you should:

• Set state column as the index.
• Create a single figure with enough Axes objects (in my example I put all of them in a single column.
• Then create each plot in corresponding Axes object.

Something like:

n = df.index.size
fig, axs = plt.subplots(n, 1, figsize=(5, 3 * n))
for i in range(n):
    df.iloc[i].plot(kind='pie', ax=axs[i], autopct='%1.1f%%')

For your data sample I got:

Of course, you can further customize particular parameters, espacially visibility of state names is to be corrected.

Or if you want to arrange your drawings in 3 columns, run:

n = df.index.size; i = 0
nRows = math.ceil(n / 3)
fig, axs = plt.subplots(nRows, 3, figsize=(14, 4 * nRows))
for ax in axs.flat:
    if i < n:
        df.iloc[i].plot(kind='pie', ax=ax, autopct='%1.1f%%')
    else:
        ax.set_visible(False)
    i += 1