try：

try:

df['group'].value_counts()['a']

切勿在熊猫中使用循环。它效率低下，通常会重新创建包中的某些现有功能。

in general, you should NEVER use for loops in pandas. it's inefficient and usually recreating some existing functionality in the package.