Skip to content

Instantly share code, notes, and snippets.

@robcarver17

robcarver17/handcrafting_test_implementation.py

Created December 14, 2018 20:43
Show Gist options
  • Star 1 You must be signed in to star a gist
  • Fork 1 You must be signed in to fork a gist
  • Save robcarver17/ff7e76a6ee74c58ae366bca9ae378ad3 to your computer and use it in GitHub Desktop.
Save robcarver17/ff7e76a6ee74c58ae366bca9ae378ad3 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
handcrafting_test_implementation.py
"""
Get some data to test the handcrafting method
"""
from sysdata.csv.csv_sim_futures_data import csvFuturesSimData
from syscore.handcrafting import Portfolio
import pandas as pd
data=csvFuturesSimData()
code_list = ['BOBL', 'BUND', 'US10', 'US20', 'KR3','KR10','EDOLLAR', 'CORN', 'CRUDE_W', 'GAS_US']
def calc_weekly_return(instrument_code, start_date=pd.datetime(2014,1,1)):
price = data[instrument_code]
price=price[start_date:]
weekly_price = price.resample("W").last()
denom_price = data.get_instrument_raw_carry_data(instrument_code).PRICE
denom_weekly_price = denom_price.resample("W").last()
weekly_returns = (weekly_price - weekly_price.shift(1))/denom_weekly_price
return weekly_returns[start_date:]
returns = dict([(instrument_code, calc_weekly_return(instrument_code)) for instrument_code in code_list])
returns = pd.DataFrame(returns)
p=Portfolio(returns, risk_target=.1)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment