CODEWITHSUNDEEP
pythonpandaslambdaiterationapply
Roman
Roman

Reputation: 105

How to use a previous row value in a pandas dataframe when the previous value is also calculated witht group data

I Have such DataFrame:

enter image description here

df = pd.DataFrame({'id': [111,111,111, 222,222,222],\
                   'Date': ['30.04.2020', '31.05.2020', '30.06.2020', \
                            '30.04.2020', '31.05.2020', '30.06.2020'],\
                   'Debt': [100,100,70, 200,200,200] , \
                   'Ear_coef': [0,0.2,0.2, 0,0,0.3]}) 
df['Date'] = pd.to_datetime(df['Date'] ) 
df['Contract'] = pd.DataFrame(df.groupby(['id']).apply(lambda x: x.Debt - x.Debt.shift(1))).reset_index().Debt
# df.groupby(['id']).
df

I need to get such DataFrame:

enter image description here

The start DataFrame:

The result DataFrame:

Ear and Debt_with_EAR at first date equels 0 and Debt at respectively.

I have tried to solve such task with apply. But I have not had a success since I need to use previous value which is also calculated. This answers do not help me Is there a way in Pandas to use previous row value in dataframe.apply when previous value is also calculated in the apply? since I Have hundreds id.

I will be grateful for the help.

Upvotes: 4

Views: 712

Answers (1)

gosuto
gosuto

Reputation: 5741

You are looking for .shift().

It does not lend itself easily for .apply() however. A work-around would be:

df['EAR'] = df['EAR_coef'] * df['Debt with EAR'].shift(1)

For you last column you might need .rolling(), but I am not sure about your formula? It seems never-ending.

Upvotes: 1

Related Questions