如何选择行中至少一个元素中包含特定值的行？

Question

I have a DataFrame DF and a list, say List1 . List1 is created from the DF and it has the elements present in DF but without repetitions. I need to do the following:
1. Select the rows of DF that contain a specific element from List1 (for instance, iterating all the elements in List1 )
2. Re-index them from 0 to whatever the number of rows are because the rows selected may have non continuous indices.

SAMPLE INPUT:

List1=['Apple','Orange','Banana','Pineapple','Pear','Tomato','Potato']
Sample DF
  EQ1      EQ2      EQ3
0 Apple    Orange   NaN
1 Banana   Potato   NaN
2 Pear     Tomato   Pineapple
3 Apple    Tomato   Pear
4 Tomato   Potato   Banana

Now if I want access to the rows that contain Apple , those would be 0 and 3. But I'd like them renamed as 0 and 1(Re-indexing). After Apple is searched, the next element from List1 should be taken and similar steps are to be carried out. I have other operations to perform after this, so I need to loop the whole process throughout List1 . I hope I have explained it well and here is my codelet for the same, which is not working:

for eq in List1:
    MCS=DF.loc[MCS_Simp_green[:] ==eq] #Indentation was missing
    MCS= MCS.reset_index(drop=True)
    <Remaining operations>

Answer 1

I think you need isin with any :

List1=['Apple','Orange','Banana','Pineapple','Pear','Tomato','Potato']

for eq in List1:
    #print df.isin([eq]).any(1)
    #print df[df.isin([eq]).any(1)]
    df1 = df[df.isin([eq]).any(1)].reset_index(drop=True)  
    print df1

     EQ1     EQ2   EQ3
0  Apple  Orange   NaN
1  Apple  Tomato  Pear
     EQ1     EQ2  EQ3
0  Apple  Orange  NaN
      EQ1     EQ2     EQ3
0  Banana  Potato     NaN
1  Tomato  Potato  Banana
    EQ1     EQ2        EQ3
0  Pear  Tomato  Pineapple
     EQ1     EQ2        EQ3
0   Pear  Tomato  Pineapple
1  Apple  Tomato       Pear
      EQ1     EQ2        EQ3
0    Pear  Tomato  Pineapple
1   Apple  Tomato       Pear
2  Tomato  Potato     Banana
      EQ1     EQ2     EQ3
0  Banana  Potato     NaN
1  Tomato  Potato  Banana

For storing values you can use dict comprehension:

dfs = {eq: df[df.isin([eq]).any(1)].reset_index(drop=True) for eq in List1}

print dfs['Apple']
     EQ1     EQ2   EQ3
0  Apple  Orange   NaN
1  Apple  Tomato  Pear

print dfs['Orange']
     EQ1     EQ2  EQ3
0  Apple  Orange  NaN

Answer 2

You can identify the items in the list and collect the resulting new DataFrame s like so:

data_frames = {}
for l in List1:
    data_frames[l] = df[df.isin([l]).any(1)].reset_index(drop=True)
    print(l, data_frames[l].index.tolist())

to get:

Apple [0, 1]
Orange [0]
Banana [0, 1]
Pineapple [0]
Pear [0, 1]
Tomato [0, 1, 2]
Potato [0, 1]

The new DataFrame objects are contained in the dictionary data_frames :

data_frames['Apple']

     EQ1     EQ2   EQ3
0  Apple  Orange   NaN
1  Apple  Tomato  Pear