对dict的操作无效

Question

我处理财务数据是为了建立一些波动性预测模型。我创建了一个名为indices的字典，其中包含我的资产

我必须在dict中的每个资产的数据帧中添加一些列。在循环之后，特别是使用indices[key]和value，我被严重卡住了

indices = dict(AEX = df[df.Symbol==".AEX"],AORD = df[df.Symbol==".AORD"],BFX = df[df.Symbol==".BFX"])

ict_items([('AEX', Unnamed: 0 Symbol    ...      Rvol    LogRvol
0     2000-01-03 00:00:00+00:00   .AEX  ...  0.000580  -7.453228
1     2000-01-04 00:00:00+00:00   .AEX  ...  5.646706   1.731072
2     2000-01-05 00:00:00+00:00   .AEX  ...  0.014028  -4.266704
3     2000-01-06 00:00:00+00:00   .AEX  ...  0.862001  -0.148499
4     2000-01-07 00:00:00+00:00   .AEX  ...  3.128320   1.140496
5     2000-01-10 00:00:00+00:00   .AEX  ...  0.129058  -2.047497
6     2000-01-11 00:00:00+00:00   .AEX  ...  0.782647  -0.245074
7     2000-01-12 00:00:00+00:00   .AEX  ...  0.271311  -1.304490
8     2000-01-13 00:00:00+00:00   .AEX  ...  0.071858  -2.633065
9     2000-01-14 00:00:00+00:00   .AEX  ...  0.727395  -0.318286

for key,value in indices.items():
    indices[key] = value.reset_index(drop = True)
    rv1 = np.full((len(value.index),1),0,dtype=float)
    rv5 = np.full((len(value.index),1),0,dtype=float)
    rv22 = np.full((len(value.index),1),0,dtype=float)
    for j in list(range(22, len(value.index), 1)):
        rv1[j] = value.iloc[j-1]['LogRvol']
        rv5[j] = np.mean(value.iloc[j - 5:j]['LogRvol'])
        rv22[j] =  np.mean(value.iloc[j - 22:j]['LogRvol'])
    indices[key] = pd.concat([value,pd.DataFrame(rv1),pd.DataFrame(rv5),pd.DataFrame(rv22)],axis=1)
    indices[key] = value.iloc[22:]
    indices[key] = value.reset_index(drop=True)

在运行代码之后，我期望得到一个更新的dict项，而实际上它们不是

这是预期的和未实现的结果

indices = dict(AEX = df[df.Symbol==".AEX"],AORD = df[df.Symbol==".AORD"],BFX = df[df.Symbol==".BFX"])

ict_items([('AEX', Unnamed: 0 Symbol    ...      Rvol    LogRvol rv1 rv5 rv22
0     2000-02-03 00:00:00+00:00   .AEX  ...  0.000580  -7.453228 
1     2000-02-04 00:00:00+00:00   .AEX  ...  5.646706   1.731072 -7.453228 
2     2000-02-05 00:00:00+00:00   .AEX  ...  0.014028  -4.266704 1.731072
3     2000-02-06 00:00:00+00:00   .AEX  ...  0.862001  -0.148499 -4.266704
4     2000-02-07 00:00:00+00:00   .AEX  ...  3.128320   1.140496 -0.148499
5     2000-02-10 00:00:00+00:00   .AEX  ...  0.129058  -2.047497 ..etc
6     2000-02-11 00:00:00+00:00   .AEX  ...  0.782647  -0.245074 
7     2000-02-12 00:00:00+00:00   .AEX  ...  0.271311  -1.304490 
8     2000-02-13 00:00:00+00:00   .AEX  ...  0.071858  -2.633065 
9     2000-02-14 00:00:00+00:00   .AEX  ...  0.727395  -0.318286 

Answer 1

示例中的for循环将indices[key]的值覆盖4次

因此，您的代码相当于：

for key, value in indices.items():
    indices[key] = value.reset_index(drop=True)

显然不会改变indices[key]的值

请注意，更新您正在迭代的项通常不是一个好主意，在本例中是indices