CODEWITHSUNDEEP
pythonpandas
Viktor.w
Viktor.w

Reputation: 2297

Pandas column turns to `True` we select two values from other columns, when `False` select the same two values

My dataframe looks like this: 

 time                    mid price  dse_high_born
18  2019-04-22 00:05:15 0.001854    False
19  2019-04-22 00:05:30 0.001854    False
20  2019-04-22 00:05:45 0.001854    False
21  2019-04-22 00:06:00 0.001854    False
22  2019-04-22 00:06:15 0.001854    True
23  2019-04-22 00:06:30 0.001854    True
24  2019-04-22 00:06:45 0.001854    True
25  2019-04-22 00:07:00 0.001856    True
26  2019-04-22 00:07:15 0.001856    True
27  2019-04-22 00:07:30 0.001856    True
28  2019-04-22 00:07:45 0.001856    True
29  2019-04-22 00:08:00 0.001856    False
30  2019-04-22 00:08:15 0.001856    False

So basically what I need is to iterate through the column dse_high_born when the column turns to True, select at the same row the value Time and mid price from the two other columns and add them into a new column Time1 and Price1 at the same index.

After that, when the column dse_high_born turns to False select the same two values (time, mid price) on the same row and add them at the same index in two new column Time2 and Price2.

I tried to iterate and use the function continue but it did not work at all... I am a bit stuck! Any idea? Thanks in advance!

The results that I need would be something like that: 

    time                    mid price  dse_high_born  Time1   price1  Time2 Price2
  2019-04-22 00:05:15 0.001854    False
  2019-04-22 00:05:30 0.001854    False
  2019-04-22 00:05:45 0.001854    False
  2019-04-22 00:06:00 0.001854    False
  2019-04-22 00:06:15 0.001854    True        2019-04-22 00:06:15 0.001854 
  2019-04-22 00:06:30 0.001854    True
  2019-04-22 00:06:45 0.001854    True
  2019-04-22 00:07:00 0.001856    True
  2019-04-22 00:07:15 0.001856    True
  2019-04-22 00:07:30 0.001856    True
  2019-04-22 00:07:45 0.001856    True                           2019-04-22 00:07:45 0.001856 
  2019-04-22 00:08:00 0.001856    False
  2019-04-22 00:08:15 0.001856    False

Upvotes: 0

Views: 39

Answers (1)

Quang Hoang
Quang Hoang

Reputation: 150785

To detect value changes, use shift():

changes = df['dse_high_born'].ne(df['dse_high_born'].shift()

and then:

changes_at_true = changes & df['dse_high_born']
changes_at_false = changes & ~df['dse_high_born']

gives changes occurs at True values. So you can do:

df.loc[changes_at_true, ['Time', 'mid_price']] += df.loc[changes_at_true, ['Time1', 'Price1']]

Edit: if you want to copy the values into new columns, do:

df['Time1'] = ''
df['Price1'] = ''

df.loc[changes_at_true, ['Time1', 'Price1']] = df.loc[changes_at_true, ['Time', 'mid_price']]

Upvotes: 1

Related Questions