finds.structured.benchmarks

Benchmarks dataset for index returns

Copyright 2022, Terence Lim

MIT License

class finds.structured.benchmarks.Benchmarks(sql: SQL, bd: BusDay, verbose: int = 0)[source]

Bases: Stocks

Provide Structured Stocks interface to benchmark and index returns

load_series(df: DataFrame, name: str, item: str = '', monthly: bool = False) → DataFrame[source]

Loads a Series containing benchmark returns to sql

Parameters:

  • df – DataFrame with time-series in each column to load to sql

  • name – Primary description to insert into ident table

  • item – Secondary description to insert into ident table

  • monthly – if True: convert index to business calendar endmo dates

Returns:

DataFrame of identifiers metadata for series successfully loaded

Notes:

  • Each column of input data frame is loaded to sql table ‘daily’, with its series name as ‘permno’ field, values as ‘ret’ field, and series index as ‘date’ field.

  • ‘idents’ table in sql is updated with identifier and descriptive info