如何更改列表中列的 dtype

Question

我正在尝试计算我有不同 dtype 的列，因为这会产生一个错误，我希望它们都具有相同的 dtype。

Question            object
Very likely         object
Quite likely       float64
Slightly likely     object
Not very likely    float64
Very unlikely       object
Total              float64
dtype: object

我尝试在 lambda 函数中使用 float(x)。

propensity = pd.read_excel(raw_data,'Propensity to buy')
propensity = propensity[['Question','Very likely','Quite likely','Slightly likely','Not very likely',
                          'Very unlikely','Total']]

propensity['MEAN'] = 
sum(
    [
        propensity['Very likely'].apply(lambda float(x): x*4),

        propensity['Quite likely'].apply(lambda float(x): x*3),

        propensity['Slightly likely'].apply(lambda float(x): x*2),

        propensity['Not very likely'].apply(lambda float(x): x*1),

        propensity['Very unlikely'].apply(lambda float(x): x*0)
    ]
)             

得到的结果是SyntaxError: invalid syntax

Answer 1

应该是

df['column_name'].apply(lambda x: float(x) * 4)

而不是

df['column_name'].apply(lambda float(x): x*4)

Answer 2

您可以通过在从源读取数据时转换列来实现这一点：
例如：

pd.read_excel('file_name', dtype={'Very likely': 'float64', 'Quite likely': 'float64', 'Slightly likely': 'float64', 'Not very likely': 'float64', 'Very unlikely': 'float64'})

Answer 3

您是否尝试过使用df['column_name'].astype(int)？如果您希望评估某些列，可以创建一个 for 循环：

cols = ['column_1','column_2','column_3','column_n']
for col in cols:
    df[col] = df[col].astype('int') 

或使用 dtype 作为条件：

if df['column_1'].dtype == object:
    print("this column contains is dtype object")
else:
    print("it is dtype is not object")

一旦将它们设置为 int 或 float，您就可以执行常规计算。