CODEWITHSUNDEEP
pandas
Ragesh Kr
Ragesh Kr

Reputation: 1723

How to retrieve previous row column value along with given column value in pandas

Below is my dataframe 

S2BillDate      totSale     count
0   2020-03-01  27880.21    796
1   2020-03-02  45664.74    1278
2   2020-03-03  41554.50    1170
3   2020-03-04  42929.59    1239
4   2020-03-05  49113.58    1380
5   2020-03-06  46945.72    1328
6   2020-03-07  12706.32    455

If my input is 2020-03-07, how can I retrieve the value above it, ie in this case 2020-03-06. Also How can i use get_value() to retrieve 'totSale' to the corresponding date ?

Is there any easier way to achieve it via pandas itself ?

Upvotes: 0

Views: 47

Answers (3)

Mayank Porwal
Mayank Porwal

Reputation: 34056

This could be a simple solution:

In [65]: df['S2BillDate'] = pd.to_datetime(df['S2BillDate'])
In [67]: df = df.set_index('S2BillDate')

In [69]: s = '2020-03-07' ## input value
In [70]: s = pd.Timestamp(s)

In [73]: loc = df.index.get_loc(s)

get_value() is now deprecated, do below instead:

In [79]: df.iloc[loc - 1]['totSale']                                                                                                                                                                        
Out[79]: 46945.72

Upvotes: 0

ansev
ansev

Reputation: 30920

We can use Series.eq and Series.shift in order to performance a boolean indexing with DataFrame.loc:

m = df['S2BillDate'].eq('2020-03-07').shift(-1, fill_value=False)

   df.loc[m]
   df.loc[m, 'S2BillDate']
   df.loc[m, ['S2BillDate', 'totSale']]

Upvotes: 1

arpitrathi
arpitrathi

Reputation: 177

To retrive the value(s) above '2020-03-07': 

df['S2BillDate'][df.S2BillDate.tolist().index('2020-03-07')-1]
>>> '2020-03-06'
df['totSale'][df.S2BillDate.tolist().index('2020-03-07')-1]
>>> 46945.72

Upvotes: 0

Related Questions