您可以使用pd.pivot_table执行此操作，如下所示：

>>> df.pivot_table(index="RepName", columns="Reason", values="Hours", fill_value=0)
Reason   Project Labor  Research  Training
RepName
John                 0         4        15
Matt                 6         0        10

完整示例：

import pandas as pd

input_data = pd.DataFrame(
    {
        "RepName": ["John", "John", "Matt", "Matt"],
        "Hours": [4, 15, 6, 10],
        "Reason": ["Research", "Training", "Project Labor", "Training"],
    }
)

output_data = pd.DataFrame(
    {
        "Project Labor": [0, 6],
        "Research": [4, 0],
        "Training": [15, 10],
    },
    index=["John", "Matt"],
)

pivoted = input_data.pivot_table(
    index="RepName",
    columns="Reason",
    values="Hours",
    fill_value=0,
)

assert output_data.equals(pivoted)

这是一个.pivot操作。透视的目的是将行的某个子集转换为它们自己的列。

out = df.pivot(index='RepName', columns='Reason', values='Hours').fillna(0)
Reason   Project Labor  Research  Training
RepName                                   
John               0.0       4.0      15.0
Matt               6.0       0.0      10.0

print(out)
Reason   Project Labor  Research  Training
RepName                                   
John               0.0       4.0      15.0
Matt               6.0       0.0      10.0

如果您不熟悉具有名称的pandas.Index，则可以通过从列索引中删除名称并重置行索引来获得更熟悉的结果：

out = (
    df.pivot(index='RepName', columns='Reason', values='Hours')
    .rename_axis(columns=None)
    .reset_index()
    .fillna(0)
)

print(out)
  RepName  Project Labor  Research  Training
0    John            0.0       4.0      15.0
1    Matt            6.0       0.0      10.0