如果您只处理一列（毛利），或者不介意将相同的格式应用于所有浮点值，下面的代码应该可以工作。如果你只需要格式化数据框中的一个列，你可能需要将值转换为字符串（因为你提到这是一个要求，所以进一步的计算会中断）。后一种情况的选择是简单地制作数据框的副本，格式化该副本以供显示，但保留原始数据以供计算。我无法让类似问题中提到的df.style.format()方法工作，可能是由于Pandas版本。

import pandas as pd
# pd version == 2.0.1

margin_data = {"margins":[0.3646, 0.2584, 0.8954]}
my_df = pd.DataFrame(margin_data)

# What you were printing before
print("Old formatting")
gross_margins = my_df['margins']
print(gross_margins)

# Apply new formatting to all floats
print("\nNew formatting")
pd.options.display.float_format = '{:.2%}'.format
gross_margins = my_df['margins']
print(gross_margins)

# Confirm values are still what you expect
print("\nValues")
print(my_df['margins'].iloc[0])

# Apply formatting to only one column, but makes 
# values a string instead of float
print("\nString formatting single column")
margin_data["more data"] = [1,2,3]
my_df2 = pd.DataFrame(margin_data)
my_df2['margins'].map('{:.2%}'.format)
print(my_df2)

生产：

Old formatting
0    0.3646
1    0.2584
2    0.8954
Name: margins, dtype: float64

New formatting
0   36.46%
1   25.84%
2   89.54%
Name: margins, dtype: float64

Values
0.3646

String formatting single column
   margins  more data
0   36.46%          1
1   25.84%          2
2   89.54%          3