Question

我有一个如下所示的CSV文件：

Build,Avg,Min,Max
BuildA,56.190,39.123,60.1039
BuildX,57.11,40.102,60.200
BuildZER,55.1134,35.129404123,60.20121

我希望得到每列的平均值，最小值，最大值，并将每个统计数据作为新行。我排除非数字列（构建列），然后运行统计信息。我通过这样做完成了这个：

df = pd.read_csv('fakedata.csv')
columns = []
builds = []

for column in df.columns:
    if(df[column].dtype == 'float64'):
        columns.append(column)
    else:
        builds.append(column)

save = df[builds]
df = df[columns]

print(df)

df.loc['Min']= df.min()
df.loc['Average']= df.mean()
df.loc['Max']= df.max()

如果我当时将这些数据写入CSV，它将如下所示：

,Avg,Min,Max
0,56.19,39.123,60.1039
1,57.11,40.102,60.2
2,55.1134,35.129404123,60.20121
Min,55.1134,35.129404123,60.1039
Average,55.8817,37.3709520615,60.1522525
Max,57.11,40.102,60.20121

哪个接近我想要的但我希望Build列再次成为第一列，并且在Min，Average，Max之上存在构建名称。基本上这个：

Builds,Avg,Min,Max
BuildA,56.19,39.123,60.1039
BuildX,57.11,40.102,60.2
BuildZER,55.1134,35.129404123,60.20121
Min,55.1134,35.129404123,60.1039
Average,55.8817,37.3709520615,60.1522525
Max,57.11,40.102,60.20121

我试图通过以下方式实现这一目标：

df.insert(0,'builds', save)
with open('fakedata.csv', 'w') as f:
    df.to_csv(f)

但这给了我这个CSV：

,builds,Avg,Min,Max
0,Build1,56.19,39.123,60.1039
1,Build2,57.11,40.102,60.2
2,Build3,55.1134,35.129404123,60.20121
Min,,55.1134,35.129404123,60.1039
Average,,55.8817,37.3709520615,60.1522525
Max,,57.11,40.102,60.20121

我该如何解决这个问题？

Answer 1

IIUC：

df_out = pd.concat([df.set_index('Build'),df.set_index('Build').agg(['max','min','mean'])]).rename(index={'max':'Max','min':'Min','mean':'Average'}).reset_index()

输出：

      index      Avg        Min       Max
0    BuildA  56.1900  39.123000  60.10390
1    BuildX  57.1100  40.102000  60.20000
2  BuildZER  55.1134  35.129404  60.20121
3       Max  57.1100  40.102000  60.20121
4       Min  55.1134  35.129404  60.10390
5   Average  56.1378  38.118135  60.16837

Python（Pandas） - 处理数字数据但返回非数字数据

1 个答案: