This message was deleted.
# hamilton-helps
Slackbot
02/07/2023, 5:50 PMThis message was deleted.
🎉 1
👀 1
s
Stefan Krawczyk
02/07/2023, 6:12 PM@Seth Stokes where should:
Copy code
INTBAL: pd.Series, ITURN: pd.Series,
YEAR: pd.Series, DAYS: pd.Seriesload_tbl_interestinterest_spread_by_monthStefan Krawczyk
02/07/2023, 6:13 PMHamilton doesn’t know how to connect the inputs required to produce a result.
Stefan Krawczyk
02/07/2023, 6:14 PMAlso with:
return load_tbl_interest.groupby([‘INTERESTMONTH’])[interest_spread].sum()Can you help me understand the intended output here? It seems like you’re referencing the
interest_spreads
Seth Stokes
02/07/2023, 6:50 PMload_tbl_interest
Seth Stokes
02/07/2023, 6:55 PMYes I didn’t know if I can reference the function or will I have to hard code the string of the column name from that function
s
Stefan Krawczyk
02/07/2023, 6:57 PMdo you mind sketching out the schema (i.e. columns) of the dataframe you want?
s
Seth Stokes
02/07/2023, 6:58 PMYes let me walk back to the office
s
Stefan Krawczyk
02/07/2023, 6:58 PMno hurry 🙂
s
Seth Stokes
02/07/2023, 7:15 PM['INTERESTMONTH', 'INTBAL', 'TYPE', ACCOUNT', 'IRATE', 'ITURN', 'YEAR', 'CCY', 'MARKET', DAYS']
Seth Stokes
02/07/2023, 7:15 PMthese are from the loaded table
s
Stefan Krawczyk
02/07/2023, 7:16 PMcool.
What should be the dataframe produced by
interest_spread_by_months
Seth Stokes
02/07/2023, 7:40 PMcorrect
Seth Stokes
02/07/2023, 7:41 PMbut also, I want the original table with the new field
interest_spreadSeth Stokes
02/07/2023, 7:45 PMI think this fixed the issue.
Copy code
@extract_columns('INTBAL', 'ITURN', 'YEAR', 'DAYS')
def load_tbl_interest(path: Union[str, List]) -> pd.DataFrame:
'''Returns data given file path.'''
if isinstance(path, str):
return pd.read_csv(path)
return pd.concat([pd.read_csv(file) for file in path])s
Stefan Krawczyk
02/07/2023, 7:46 PMyep that is one way to expose it
s
Seth Stokes
02/07/2023, 7:47 PMbut there are new issues, so please ignore what I had said
s
Stefan Krawczyk
02/07/2023, 7:47 PM Copy code
df = dr.execute(['interest_spread'])s
Seth Stokes
02/07/2023, 7:49 PMHow can the new column
interest_spreadload_tbl_interestload_tbl_interestSeth Stokes
02/07/2023, 7:50 PMWould I have to either create a new funciton with a join? Or just use dataframes instead and return the resulting dataframe?
s
Stefan Krawczyk
02/07/2023, 7:51 PMOne way:
In your hamilton functions
Copy code
@extract_columns(*['INTERESTMONTH', 'INTBAL', 'TYPE', ACCOUNT', 'IRATE', 'ITURN', 'YEAR', 'CCY', 'MARKET', DAYS'])
def load_tbl_interest(path: Union[str, List]) -> pd.DataFrame:
... Copy code
df = dr.execute(['INTERESTMONTH', 'INTBAL', 'TYPE', ACCOUNT', 'IRATE', 'ITURN', 'YEAR', 'CCY', 'MARKET', DAYS', 'interest_spread'])👍 1
Stefan Krawczyk
02/07/2023, 7:52 PMand hamilton will join it
Stefan Krawczyk
02/07/2023, 7:52 PMor do it explicitly in a function as you suggested.
s
Seth Stokes
02/07/2023, 7:57 PMbeautiful, thank you sir. so given that I will usually always want all the columns i should just decorate the load functions and then ask for all beginning columns and functions in the driver
👍 1
3 Views