CODEWITHSUNDEEP
pythonpandas
user266003
user266003

Reputation:

Search by row in pandas

I have an hdf file which is printed out like this:

print dt['data1']
=>

              column1  column2  column3  
row1                3        8        2        
row2                4        3        0  
row3                12       1        14
row4                -5       5        25

1) How can I search by rows or rather "select" it? That is, 

dt.get_row('row3')
=> row3                12       1        14

2) And how can I search by row and column value? That is 

dt.get_row('row2').where(column_value=3)
=> column2

Upvotes: 1

Views: 3989

Answers (2)

Mike Müller
Mike Müller

Reputation: 85442

Setting your the pandas DataFrame to df:

df = dt['data1']

You can get the row with ix:

 df.ix['row3']

This gives the result as a Series:

column1    12
column2     1
column3    14
Name: row3, dtype: int64

You can also access the values as NumPy a array:

>>> df.ix['row3'].values
array([12,  1, 14])

or the name of the row:

>>> df.ix['row3'].name
'row3'

and the column in row2 where the value is equal to 3 with:

r2 = df.ix['row2']
r2[r2.eq(3)].index

This gives you an Index object:

Index(['column2'], dtype='object')

You can access single elements too:

>>> r2[r2.eq(3)].index[0]
'column2'

Upvotes: 1

Bharath M.S
Bharath M.S

Reputation: 31

Here is a code snippet that I could get working. Not sure if it is the best way to do it.

>>> import pandas as pd
>>> df = pd.DataFrame({'AAA' : [4,5,6,7], 'BBB' : [10,20,30,40],'CCC' : [100,50,-30,-50]})
>>> df
    AAA  BBB  CCC<br>
0    4   10  100<br>
1    5   20   50<br>
2    6   30  -30<br>
3    7   40  -50<br>
>>> df.iloc[0]
AAA      4
BBB     10
CCC    100
Name: 0, dtype: int64
>>> df.iloc[0].AAA
4

Hope this is what you were looking for.

Upvotes: 0

Related Questions