Hi - I am trying to run a minimum variance portfolio optimization and scipy.optimize.minimize keeps giving me a very generic error. Can anybody help me out? Debug shows it errors out when executing the minimize() function. This code runs perfectly offline when I replicate the price data (using Python 3.4 and scipy 0.14.0)

import datetime  
from pandas import Timestamp  
import numpy as np  
import pandas as pd  
import math  
from scipy.optimize import minimize

def initialize(context):  
    context.secs = [  
                    sid(8554),  
                    sid(21769),  
                    sid(14520),  
                    sid(24705),  
                    sid(26669),  
                    sid(33072),  
                    sid(23870),  
                    sid(23921),  
                    sid(28054),  
                    sid(26981)  
                   ]

    context.cov_lookback = 60

    set_commission(commission.PerTrade(cost=0.05))  
    schedule_function(  
        func=rebalance,  
        date_rule=date_rules.week_start(days_offset=1),  
        time_rule=time_rules.market_open(minutes=1)  
    )

def meanvaropt(x, cov_matrix):  
    scaling = 1/x.sum()  
    x = x * scaling  
    port_var_components = cov_matrix * x  
    port_var_components = port_var_components.transpose() * x  
    port_var = port_var_components.sum()  
    return port_var

def rebalance(context, data):  
    df = history(252, '1d', 'price').filter(items = context.secs)  
    securities_df = pd.DataFrame(data=np.ones(len(df.columns)),index=df.columns.values,columns=['weight'])

    # calculate covariance matrix for selected stocks  
    selected_returns = df.filter(items = securities_df.index).pct_change()[1:]  
    selected_returns_cov = selected_returns[len(selected_returns) - context.cov_lookback:].cov()  
    # solve minimum variance optimization for portfolio of selected securities  
    solution = minimize(meanvaropt,  
                        securities_df['weight'].values,  
                        args = [selected_returns_cov.as_matrix()],  
                        method = 'L-BFGS-B',  
                        bounds = [(0.,1.) for _ in range(len(securities_df['weight']))],  
                        options = {'disp':True}  
                        )  
    scaling = 1 / solution.x.sum()  
    securities_df['weight'] = solution.x * scaling

    for sec in securities_df.index:  
        order_target_percent(sec, securities_df['weight'][sec])  
def handle_data(context, data):  
    return