Generate a DataFrame with one row per benchmark. A benchmark is any object that has a function valuation_details(...) that returns a dictionary of the right shape. Allowed benchmarks at the moment are depos, fras and swaps. Various useful information is reported. This is a bit slow so do not use in performance-critical spots

Crete IborBenchmarks from a dataframe. The dataframe should have at least these columns with these sample inputs: type start_dt maturity_dt dc_type notional contract_rate fixed_leg_type fixed_freq_type 0 IborDeposit 06-OCT-2001 09-OCT-2001 ACT_360 100.0 0.042 NaN NaN 1 IborFRA 09-JAN-2002 09-APR-2002 ACT_360 100.0 0.042 PAY NaN 2 IborSwap 09-OCT-2001 09-OCT-2002 THIRTY_E_360_ISDA 1000000 0.042 PAY SEMI_ANNUAL start_dt and maturity_dt could be Date, string convertible to Tenor, or datetime Args: df (pd.DataFrame): dataframe as bove, with benchmark info asof_date (Date): if start_dt is a tenor string, asof_date is used as an anchor for that calendar_type (CalendarTypes): What calendar to use Returns: dict: Keys are benchmark types as in df['type'], values are lists of benchmarks of that type

Generate a DataFrame with one row per bechmark used in constructing a given ibor_curve. Various useful information is reported. This is a bit slow so do not use in performance-critical spots