使用数据框仅从字典中选择所需的键

Question

我有一个数据框，其中包含产品及其状态，如下所示

数据帧：

products    status
11  sale
22  sale
33  notsale
44  notsale
55  notsale
66  removed
77  removed
88  notsale
99  sale
222 sale
333 removed
444 removed
555 notsale

我还将用户数据作为带有用户的字典和他们感兴趣的产品列表。

{1: [11,22,33,555,33], 2:[33,66,77,88,99],3:[11,88,99,222,333,555],4:[333,33,444,44],5:[333,444,22,33,44,55,66]}

我需要做的是，删除状态为removed的产品以及用户对上述字典感兴趣的重复项。

预期产量：

{1: [11,22,33,555,], 2: [33, 88,99], 3:[11,88,99,222,555], 4: [33, 44], 5: [22, 33,44,55]}

Answer 1

首先使用boolean indexing对removed值进行过滤，然后在dict comprehension中将值转换为set以获取唯一值，然后删除a的值：

a = df.loc[df['status'] == 'removed', 'products'].tolist()
print (a)
[66, 77, 333, 444]

d = {1: [11,22,33,555,33], 2:[33,66,77,88,99], 
     3:[11,88,99,222,333,555], 4:[333,33,444,44],5:[333,444,22,33,44,55,66]}

d1 = {k: list(set(v)-set(a)) for k, v in d.items()}
print (d1)
{1: [33, 11, 22, 555], 2: [88, 33, 99], 
 3: [11, 555, 99, 222, 88], 4: [33, 44], 5: [33, 44, 22, 55]}

编辑：

使用qazxsw poi的多个关键字过滤：

isin