CODEWITHSUNDEEP
pythonpandasdataframe
Snowfire777
Snowfire777

Reputation: 203

Browsing and selecting on 2 dataframes

Let's say I have 2 DataFrames like this:

df1
     Id     B
3    alpha  1
18   alpha  3
125  alpha  5
230  alpha  9

where 3, 18, 125, 230 are the index of df1

and: 

df2
     Id     B
1    Beta  21
2    Beta  33
5    Beta  120
7    Beta  36
10   Beta  32
14   Beta  71
15   Beta  210
21   Beta  53
123  Beta  22 
127  Beta  1227
128  Beta  11
227  Beta  7
235  Beta  18

My question : How can I browse my df2 to extract the rows coming before and after each row of my df1 based on the index?

My desired output :

df2
     Id     B
2    Beta  33
5    Beta  120
15   Beta  210
21   Beta  53
123  Beta  22 
127  Beta  1227
227  Beta  7
235  Beta  18

Upvotes: 2

Views: 40

Answers (2)

iDrwish
iDrwish

Reputation: 3103

A more verbose way to accomplish this is get the location of the indexes from df1 and plus-minus one to them.

df = pd.concat([df1, df2]).sort_index()

li = []
for ind in df1.index:
    ind_loc = df.index.get_loc(ind)
    li.extend([ind_loc - 1, ind_loc + 1])

print(df.iloc[li])

Upvotes: 1

jezrael
jezrael

Reputation: 863226

Use searchsorted for positions, add one previous values, join by numpy.r_ and select by iloc:

a = df2.index.searchsorted(df1.index) 
df = df2.iloc[sorted(np.r_[a, a-1])]
print (df)
       Id     B
2    Beta    33
5    Beta   120
15   Beta   210
21   Beta    53
123  Beta    22
127  Beta  1227
227  Beta     7
235  Beta    18

Upvotes: 3

Related Questions