The error is saying 'str' object has no attribute 'end_date'. This is because the order_optimal_portfolio method is being passed a series of target weights which has its index set to the stock tickers (ie strings). The order_optimal_portfolio expects the index to be asset objects and not strings. This is true for just about every method on Quantopian. Whenever referencing assets one needs to reference the asset objects and not simply the ticker. The way to convert tickers to asset objects, in an algo, is with one of the three methods symbol, symbols, sid.

So, a fix could be to change line 36 from

    context.symbol_list = [VTI', 'TLT', 'IEF', 'GLD', 'DBC']

to

    context.symbol_list = symbols('VTI', 'TLT', 'IEF', 'GLD', 'DBC')

That will create a list of asset objects from the tickers. You could do other things but that is an immediate fix. Good luck.

Here is my simple way to do the same without pipeline.

import quantopian.optimize as opt  
import pandas as pd  
# ------------------------------------------------------------------------------------------  
assets = symbols('VTI', 'TLT', 'IEF', 'GLD', 'DBC'); weights = [1.0, 0, 0, 0, 0]; lev = 1.0;  
# ------------------------------------------------------------------------------------------  
def initialize(context):  
     schedule_function(rebalance, date_rules.every_day(), time_rules.market_open(minutes = 65))         

def rebalance(context,data):  
    assets_to_trade_with_weights = pd.Series(index = assets, data = weights)

    order_optimal_portfolio(opt.TargetWeights(assets_to_trade_with_weights), [opt.MaxGrossExposure(lev)])  

@Dan Whitnable,
please respond to my request.

Thanks Dan and Vladimir! Thats very helpful. I have managed to get the algorithm working as expected.

Take care

Ojas