使用来自 dataframe2 的值过滤 dataframe1 并在 Python 中的特定行值之后选择 dataframe 1 中的所有行

Question

我有 2 个数据帧（df1 &df2）。我想从 df2 中获取每一行，对于特定的 ID 和查询，在找到“查询”字符串后从 df1 中获取该特定 ID 的所有值。唯一的条件是显示的行的“发送者”列值应仅为“支持团队”。

我尝试过类似df1= df1.loc['can we find tigers in Amazon forest?':] 但是得到了keyerror..任何人都可以帮我解决这个问题..

Note:Index in df1 is not sorted as the dataframes are grouped based on ID

df1 =

Index  ID        Query                                    Sent by
0    76649  Hi                                           Jack
2    76649  Anyone there                                 Jack
3    76649  yes hi                                    Support Team
10   76649  this is Fred from support team            Support Team
5    76649  can we find tigers in Amazon forest?        Jack
6    76649  yes tigers can be found there             Support Team
7    76649  contact forest dept for more              Support Team
13   76649  thanks for reaching out                   Support Team
9    67209  Hello                                      Bianca
4    67209  Anyone there                              Bianca
11   67209  Hi this is Jim from support team          Support Team
12   67209  can we find lions in Amazon forest?       Bianca
8    67209  yes lions  can be found there             Support Team
14   67209  contact forest dept for more              Support Team
15   67209  thanks for reaching out                   Support Team
16   67209  sure that helps..thank you                Bianca

df2 =

Index        Query                                         ID
0      can we find tigers in Amazon forest?               76649
2      can we find lions in Amazon forest?                67209
3      can we find elephant in Amazon forest?             77832

预期输出：

76649  yes tigers can be found there             Support Team
76649  contact forest dept for more              Support Team
76649  thanks for reaching out                   Support Team
67209   yes lions  can be found there             Support Team
67209   contact forest dept for more              Support Team
67209   thanks for reaching out                   Support Team

Answer 1

我不知道这是否是最优雅的方式。但这将是我的尝试。如果有不清楚的地方，请要求澄清。

# Just example dfs for testing
df1 = pd.DataFrame({'Query': ['this is Fred from support team', 'can we find tigers in Amazon forest?', 'yes tigers can be found there', 'can we find lions in Amazon forest?','yes lions can be found there'],
                    'ID':   [1,1,1,2,2]})
df2 = pd.DataFrame({'Query': ['can we find tigers in Amazon forest?', 'can we find lions in Amazon forest?'],
                    'ID':   [1,2]})

# Reset index so we can take every object with greater index
df1.reset_index(inplace=True)

#init output
output = None

#iterate over df2
for idx, row in df2.iterrows():

    # Find index of matching string and id in df1
    index = df1.index[(df1['Query'] == row['Query']) & (df1['ID'] == row['ID'])]

    # index is a list so check if the result is consistent with our logic
    # If string not found
    if len(index) == 0:
        continue

    # You can add here code what to do if the same string with same id appears more often in df1
    elif len(index) > 1:
        print("Oopsi, your string seems to appear more often with the same ID!")

    else:

        # Create output
        if output is None:
            output = df1[(df1['ID'] == row['ID']) & (df1.index > index[0])]
        else:
            output = output.append(df1[(df1['ID'] == row['ID']) & (df1.index > index[0])])

# Filter by support team
output = output[output['Sent by'] == 'Support Team']
print(output)