Panda .loc或.iloc从数据集

Question

我一直在尝试从数据集中为所有行选择一组特定的列。我尝试过类似下面的内容。

train_features = train_df.loc[,[0,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18]]

我想提一下所有行都包含但只需要编号的列。 有没有更好的方法来解决这个问题。

示例数据：

age  job        marital   education    default   housing   loan   equities   contact     duration   campaign   pdays   previous   poutcome   emp.var.rate   cons.price.idx   cons.conf.idx   euribor3m     nr.employed   y
56   housemaid  married   basic.4y     1         1         1      1          0           261        1          999     0          2          1.1            93.994           -36.4           3.299552287   5191          1
37   services   married   high.school  1         0         1      1          0           226        1          999     0          2          1.1            93.994           -36.4           0.743751247   5191          1
56   services   married   high.school  1         1         0      1          0           307        1          999     0          2          1.1            93.994           -36.4           1.28265179    5191          1

我试图忽略我的数据集中的工作，婚姻，教育和y列。 y列是目标变量。

Answer 1

如果需要按职位选择，请使用iloc：

train_features = train_df.iloc[:, [0,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18]]
print (train_features)
   age  default  housing  loan  equities  contact  duration  campaign  pdays  \
0   56        1        1     1         1        0       261         1    999   
1   37        1        0     1         1        0       226         1    999   
2   56        1        1     0         1        0       307         1    999   

   previous  poutcome  emp.var.rate  cons.price.idx  cons.conf.idx  euribor3m  \
0         0         2           1.1          93.994          -36.4   3.299552   
1         0         2           1.1          93.994          -36.4   0.743751   
2         0         2           1.1          93.994          -36.4   1.282652   

   nr.employed  
0         5191  
1         5191  
2         5191  

另一个解决方案是drop不必要的列：

cols= ['job','marital','education','y']
train_features = train_df.drop(cols, axis=1)
print (train_features)
   age  default  housing  loan  equities  contact  duration  campaign  pdays  \
0   56        1        1     1         1        0       261         1    999   
1   37        1        0     1         1        0       226         1    999   
2   56        1        1     0         1        0       307         1    999   

   previous  poutcome  emp.var.rate  cons.price.idx  cons.conf.idx  euribor3m  \
0         0         2           1.1          93.994          -36.4   3.299552   
1         0         2           1.1          93.994          -36.4   0.743751   
2         0         2           1.1          93.994          -36.4   1.282652   

   nr.employed  
0         5191  
1         5191  
2         5191  

Answer 2

您可以通过底层的numpy数组

访问列值

考虑数据框df

df = pd.DataFrame(np.random.randint(10, size=(5, 20)))
df

您可以切片底层数组

slc = [0,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18]
df.values[:, slc]

array([[1, 3, 9, 8, 3, 2, 1, 6, 6, 0, 3, 9, 8, 5, 9, 9],
       [8, 0, 2, 3, 7, 8, 9, 2, 7, 2, 1, 3, 2, 5, 4, 9],
       [1, 1, 9, 3, 5, 8, 8, 8, 8, 4, 8, 0, 5, 4, 9, 0],
       [6, 3, 1, 8, 0, 3, 7, 9, 9, 0, 9, 7, 6, 1, 4, 8],
       [3, 2, 3, 3, 9, 8, 3, 8, 3, 4, 1, 6, 4, 1, 6, 4]])

或者您可以从此切片重建新数据框

pd.DataFrame(df.values[:, slc], df.index, df.columns[slc])

不与

一样干净直观

df.iloc[:, slc]

您还可以使用slc对df.columns对象进行切片并将其传递给df.loc

df.loc[:, df.columns[slc]]