如何根据基于另一个数据帧的条件提取熊猫数据帧的行

Question

I have these two dataframes :

df1 = pd.DataFrame({'Points':[1,2,3,4,5], 'ColX':[9,8,7,6,5]})
df1
    Points  ColX
0        1     9
1        2     8
2        3     7
3        4     6
4        5     5

df2 = pd.DataFrame({'Points':[2,5], 'Sum':[-1,1], 'ColY':[2,4]}) # ColY does not matter, I just added it to say that this dataframe can have other columns that the useful columns for this topic
df2
    Points  Sum  ColY
0        2   -1     2
1        5    1     4

I would like to get a dataframe with the rows of df1 where :

• the values of column Points in df1 are also in the the column Points of df2
• the values of column Sum in df2 are between 0 and 2

Consequently, I would like to get this dataframe (no matter the index) :

    Points  ColX
4        5     5

I tried the following but it didn't work :

df1[df1.merge(df2, on = 'Points')['Sum'] <= 2 and ['Sum']>=0] 

Could you please help me to find the right code ?

Answer 1

Try this:

df1[df1['Points'].isin(df2.query('0 <= Sum <= 2')['Points'])]

Output:

  Points  ColX
4       5     5

Explained:

• df2.query('0 <= Sum <=2') to filter df2 first to only valid records
• Then use boolean indexing with isin of filter df2 Points column.

Answer 2

Use Series.between for boolean mask with boolean indexing for filtering passed to another mask with Series.isin :

df = df1[df1['Points'].isin(df2.loc[df2['Sum'].between(0,2), 'Points'])]
print (df)
   Points  ColX
4       5     5

Your solution should be changed with DataFrame.query for filtering:

df = df1.merge(df2, on = 'Points').query('0<=Sum<=2')[df1.columns]
print (df)
   Points  ColX
1       5     5

Answer 3

also works:

df3 = df1.merge(df2, on='Points')
result = df3[(df3.Sum >= 0) & (df3.Sum <= 2)]
result