python - 如何获取 Pandas 数据框中列的百分比？

Question

我正在关注this article这是使用 Pandas。

作者在其中指出:

Most traffic stops are of white drivers, which is to be expected since Vermont is around 94% white (making it the 2nd-least diverse state in the nation, behind Maine). Since white drivers make up approximately 94% of the traffic stops, there's no obvious bias here for pulling over non-white drivers vs white drivers. Using the same methodology, however, we can also see that while black drivers make up roughly 2% of all traffic stops, only 1.3% of Vermont's population is black.

但他没有展示如何使用 Pandas 来计算这一点。

这是数据的示例

In[165: df_vt['driver_race'].value_counts()
Out[15]:
White       261339
Black         5571
Asian         3446
Hispanic      2562
Other          263
Name: driver_race, dtype: int64

我发现我可以这样做:

df_vt.groupby(('driver_race')).size() / len(df_vt)

这会给我:

driver_race
Asian       0.012614
Black       0.020393
Hispanic    0.009378
Other       0.000963
White       0.956651
dtype: float64

这几乎是我想要的(它确实给了我 2%)，但我很想看到类似的东西

 White     261339     0.956651
 Black       5571     0.020393
 .
 .
 .
 Other        263     0.000963

如何使用 Pandas 获得这个？

Answer 1

这非常简单。计算 value_counts 并将其转换为数据帧。然后，只需计算百分比列即可。

vc = df_vt['driver_race'].value_counts().to_frame('counts')
vc['%'] = vc['counts'] / vc['counts'].sum()

vc

          counts         %
White     261339  0.956651
Black       5571  0.020393
Asian       3446  0.012614
Hispanic    2562  0.009378
Other        263  0.000963