I've a dateset like this:

I want a new column date_new which should look at the condition in next column. If condition is one, do nothing. If condition is 2, add a day to the date and store in date_new. Additional condition- There should be 3 continuous 2's for this to work.

The final output should look like this.

Any help is appreciated. Thank you.

CodePudding user response：

This solution is a little bit different. If condition is 1 put None, otherwise I add condition value -1 to the date

df['date_new'] = np.where(df['condition'] == 1, None, (df['date']   pd.to_timedelta(df['condition']-1,'d')).dt.strftime('%d-%m-%Y') )

CodePudding user response：

Ok, so I've edited my answer and transform it into a function:

def newdate(df):
 L = df.Condition
 res = [i for i, j, k in zip(L, L[1:], L[2:]) if i == j == k]
 if 2 in res:
  df['date'] = pd.to_datetime(df['date'])
  df['new_date'] = df.apply(lambda x: x["date"] pd.DateOffset(days=2) if x["Condition"]==2 else pd.NA, axis=1)
  df['new_date'] = pd.to_datetime(df['new_date'])
  df1 = df
  return df1

#output: