AttributeError: 'str' object has no attribute '_name' - The error is clear - but for the life of me I can't see why it's happening

[Capn_Hacky]

It's a multi-data stratergy and I have data objects in the 'next' function cycling as the entries in self.datas - in this function everything seems to work fine in interating with the data objects, in partcular they certainly do have name attributes - this error seems to occur after the cycle of "next's" is complete.

Apologies if this turns out to be something trivial.

The stack trace is as follows:

Traceback (most recent call last):
  File "trader.py", line 472, in <module>
    runstrategy(stocks_of_interest,financial_history, pairs_list)
  File "trader.py", line 346, in runstrategy
    oldsync=args.oldsync)
  File "C:\ProgramData\Anaconda3\envs\finance\lib\site-packages\backtrader\cerebro.py", line 1127, in run
    runstrat = self.runstrategies(iterstrat)
  File "C:\ProgramData\Anaconda3\envs\finance\lib\site-packages\backtrader\cerebro.py", line 1293, in runstrategies
    self._runonce(runstrats)
  File "C:\ProgramData\Anaconda3\envs\finance\lib\site-packages\backtrader\cerebro.py", line 1688, in _runonce
    self._brokernotify()
  File "C:\ProgramData\Anaconda3\envs\finance\lib\site-packages\backtrader\cerebro.py", line 1360, in _brokernotify
    self._broker.next()
  File "C:\ProgramData\Anaconda3\envs\finance\lib\site-packages\backtrader\brokers\bbroker.py", line 1240, in next
    self._get_value()  # update value
  File "C:\ProgramData\Anaconda3\envs\finance\lib\site-packages\backtrader\brokers\bbroker.py", line 440, in _get_value
    comminfo = self.getcommissioninfo(data)
  File "C:\ProgramData\Anaconda3\envs\finance\lib\site-packages\backtrader\broker.py", line 80, in getcommissioninfo
    if data._name in self.comminfo:
AttributeError: 'str' object has no attribute '_name'

For clarity I can attach the code if that would be helpful but I didn't want to clutter this post unnecessarily?

Asked for code:

class PairTradingStrategy(bt.Strategy):
    params = dict(
        period=300,
        stake=1,
        qty1=0,
        qty2=0,
        printout=True,
        upper=1,
        lower=-1,
        up_medium=0.75,
        low_medium=-0.75,
        status=0,
        portfolio_value=50,
    )

    def log(self, txt, dt=None):
        if self.p.printout:
            dt = dt or self.data.datetime[0]
            dt = bt.num2date(dt)
            print('%s, %s' % (dt.isoformat(), txt))

    def notify_order(self, order):
        if order.status in [bt.Order.Submitted, bt.Order.Accepted]:
            return  # Await further notifications

        if order.status == order.Completed:
            if order.isbuy():
                buytxt = 'BUY COMPLETE, %.2f' % order.executed.price
                self.log(buytxt, order.executed.dt)
            else:
                selltxt = 'SELL COMPLETE, %.2f' % order.executed.price
                self.log(selltxt, order.executed.dt)

        elif order.status in [order.Expired, order.Canceled, order.Margin]:
            self.log('%s ,' % order.Status[order.status])
            pass  # Simply log

        # Allow new orders
        self.orderid = None

    def __init__(self,pairs_list):
        # To control operation entries
        self.pairs_list = pairs_list
        self.orderid = None
        self.qty1 = self.p.qty1
        self.qty2 = self.p.qty2
        self.upper_limit = self.p.upper
        self.lower_limit = self.p.lower
        self.up_medium = self.p.up_medium
        self.low_medium = self.p.low_medium
        self.status = self.p.status
        self.portfolio_value = self.p.portfolio_value


        #kalman filter stuff -- curently not sure what its about, just want that beta
        self.delta = 0.0001
        self.Vw = self.delta / (1 - self.delta) * np.eye(2)
        self.Ve = 0.001
        self.beta = np.zeros(2)
        self.P = np.zeros((2, 2))
        self.R = np.zeros((2, 2))


        # Signals performed with PD.OLS :
        self.transform = btind.OLS_TransformationN(self.data0, self.data1,
                                                   period=self.p.period)
        self.zscore = self.transform.zscore
        #self.beta = pd.ols(self.data0, self.data1, window_type=self.p.period)


        x = np.asarray([self.data0[0], 1.0]).reshape((1, 2))
        y = self.data1[0]

        self.R = self.P + self.Vw  # state covariance prediction
        yhat = x.dot(self.beta)  # measurement prediction

        Q = x.dot(self.R).dot(x.T) + self.Ve  # measurement variance

        e = y - yhat  # measurement prediction error

        K = self.R.dot(x.T) / Q  # Kalman gain

        self.beta += K.flatten() * e  # State update

        self.P = self.R - K * x.dot(self.R)

        sqrt_Q = np.sqrt(Q)

        self.lower_limit = -1.5*sqrt_Q
        self.upper_limit = 1.5*sqrt_Q

        self.up_medium = sqrt_Q
        self.low_medium = -sqrt_Q

        # Checking signals built with StatsModel.API :
        # self.ols_transfo = btind.OLS_Transformation(self.data0, self.data1,
        #                                             period=self.p.period,
        #                                             plot=True)

    def next(self):

        ########PRETTTYYYYYYYYYY SURE SHORT SELLS ARENT WORKING AS THEY SHOULDDDD......

        # data is in self.datas

        # for i, d in enumerate(self.datas):
        #     dt, dn = self.datetime.date(), d._name
        #     pos = self.getposition(d).size

        #     for p in pairs_list:
        #         if p[0] == d._name:
        #             #no current market orders
        #             if pos == None

            # if self.orderid:
            #     return  # if an order is active, no new orders are allowed

        def get_data_by_name(name):
            output = None
            for data in self.datas:
                #print(data)
                if data._name == name:
                    output = data


            if output is None:
                #shoulf probably create a log
                print('Warning: The data with name: '+ name+ ' could not be found.')

            return(output)

        print(self.pairs_list)

        for pair in self.pairs_list:
            p0_data = get_data_by_name(pair[0])
            p1_data = get_data_by_name(pair[1])
            print(pair)


            #if data isnt available just skip it
            if p0_data is not None and p1_data is not None :
                print(p0_data)
                print(p1_data)
                print(p0_data._name)
                print(p1_data._name)

                print('positions')
                print(self.getposition(pair[0]) )
                print(self.getposition(pair[1]) )

                if self.getposition(pair[0]).size ==0 and self.getposition(pair[1]).size == 0:

                    print('hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh')
                    if self.p.printout:
                        print('Self  len:', len(self))
                        print('Data0 len:', len(self.data0))
                        print('Data1 len:', len(self.data1))
                        print('Data0 len == Data1 len:',
                              len(p0_data) == len(p1_data))

                        print('Data0 dt:', p0_data.datetime.datetime())
                        print('Data1 dt:', p1_data.datetime.datetime())

                    print('status is', self.status)
                    print('zscore is', self.zscore[0])

                    # Step 2: Check conditions for SHORT & place the order
                    # Checking the condition for SHORT
                    if (self.zscore[0] > self.upper_limit) and (self.status != 1):

                        # Calculating the number of shares for each stock
                        value = 0.5 * self.portfolio_value  # Divide the cash equally
                        x = int(value / (p0_data.close))  # Find the number of shares for Stock1
                        y = int(value / (p1_data.close))  # Find the number of shares for Stock2
                        print('x + self.qty1 is', x + self.qty1)
                        print('y + self.qty2 is', y + self.qty2)

                        # Placing the order
                        self.log('SELL CREATE %s, price = %.2f, qty = %d' % (p0_data._name, p0_data.close[0], x + self.qty1))
                        self.sell(data=p0_data, size=(x + self.qty1))  # Place an order for buying y + qty2 shares
                        self.log('BUY CREATE %s, price = %.2f, qty = %d' % (p1_data._name, p1_data.close[0], y + self.qty2*self.beta[0]))
                        self.buy(data=p1_data, size=(y + self.qty2*self.beta[0]))  # Place an order for selling x + qty1 shares

                        # Updating the counters with new value
                        self.qty1 = x  # The new open position quantity for Stock1 is x shares
                        self.qty2 = y  # The new open position quantity for Stock2 is y shares

                        # doesnt make sense for multiple feeds self.status = 1  # The current status is "short the spread"

                        # Step 3: Check conditions for LONG & place the order
                        # Checking the condition for LONG
                    elif (self.zscore[0] < self.lower_limit) and (self.status != 2):

                        # Calculating the number of shares for each stock
                        value = 0.5 * self.portfolio_value  # Divide the cash equally
                        x = int(value / (self.p0_data.close))  # Find the number of shares for Stock1
                        y = int(value / (self.data1.close))  # Find the number of shares for Stock2
                        print('x + self.qty1 is', x + self.qty1)
                        print('y + self.qty2 is', y + self.qty2)

                        # Place the order
                        self.log('BUY CREATE %s, price = %.2f, qty = %d' % (p0_data._name, p0_data.close[0], x + self.qty1))
                        self.buy(data=p0_data, size=(x + self.qty1))  # Place an order for buying x + qty1 shares
                        self.log('SELL CREATE %s, price = %.2f, qty = %d' % (p1_data._name, p1_data.close[0], y + self.qty2*self.beta[0]))
                        self.sell(data=p1_data, size=(y + self.qty2*self.beta[0]))  # Place an order for selling y + qty2 shares

                        # Updating the counters with new value
                        self.qty1 = x  # The new open position quantity for Stock1 is x shares
                        self.qty2 = y  # The new open position quantity for Stock2 is y shares
                        self.status = 2  # The current status is "long the spread"


                        # Step 4: Check conditions for No Trade
                        # If the z-score is within the two bounds, close all
                
                elif (self.zscore[0] < self.up_medium and self.zscore[0] > self.low_medium):
                    self.log('CLOSE LONG %s, price = %.2f' % (p0_data._name, p0_data.close[0]))
                    self.close(p0_data)
                    self.log('CLOSE LONG %s, price = %.2f' % (p1_data._name, p1_data.close[0]))
                    self.close(p1_data)
                print('yyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyy')

            print('got to the ennnnnnnnnndddddddddddddddddd')    

    def stop(self):
        print('==================================================')
        print('Starting Value - %.2f' % self.broker.startingcash)
        print('Ending   Value - %.2f' % self.broker.getvalue())
        print('==================================================')





def runstrategy(stocks_of_interest,financial_history_dict, pairs_list):

    args = parse_args()

    # Create a cerebro
    cerebro = bt.Cerebro()

    # Get the dates from the args
    fromdate = datetime.datetime.strptime(args.fromdate, '%Y-%m-%d')
    todate = datetime.datetime.strptime(args.todate, '%Y-%m-%d')

    # # Create the 1st data
    # data0 = btfeeds.YahooFinanceCSVData(
    #     dataname=args.data0,
    #     fromdate=fromdate,
    #     todate=todate)

    # # Add the 1st data to cerebro
    # cerebro.adddata(data0)

    # # Create the 2nd data
    # data1 = btfeeds.YahooFinanceCSVData(
    #     dataname=args.data1,
    #     fromdate=fromdate,
    #     todate=todate)

    # # Add the 2nd data to cerebro
    # cerebro.adddata(data1)

    ############################# data other 


    # # Simulate the header row isn't there if noheaders requested
    # skiprows = 1 if args.noheaders else 0
    # header = None if args.noheaders else 0

    #must be a better way to do this
    stocks_in_order_used = []
    skipped = []
    for stock in stocks_of_interest:
        try:
            dataframe = financial_history_dict[stock]
            data = bt.feeds.PandasData(dataname=dataframe)
            cerebro.adddata(data, name=stock)
        except KeyError:
                print('skipping ' + stock)
                skipped += [stock]
        
        #print(dataframe)


    #bcause doing it here prevents things fucking up
    adj_pairs_list=[]
    print(pairs_list)
    for pair in pairs_list:
        dont_skip = True
        for skip_stock in skipped:
            if skip_stock in pair:
                dont_skip = False


        if dont_skip:
            adj_pairs_list.append(pair) 

    print('adjjjjjjjjjjjjjjj')
    print(adj_pairs_list)

    # Add the strategy
    cerebro.addstrategy(PairTradingStrategy,pairs_list = adj_pairs_list,
                        period=args.period,
                        stake=args.stake)

    # Add the commission - only stocks like a for each operation
    cerebro.broker.setcash(args.cash)

    # Add the commission - only stocks like a for each operation
    cerebro.broker.setcommission(commission=args.commperc)

    # And run it
    #cerebro.run()


    cerebro.run(runonce= not args.runnext,
                 preload= not args.nopreload,
                 oldsync=args.oldsync)

    # Plot if requested
    if args.plot:
        cerebro.plot(numfigs=args.numfigs, volume=False, zdown=False)


def parse_args():
    parser = argparse.ArgumentParser(description='MultiData Strategy')

    parser.add_argument('--data0', '-d0',
                        default='data/csvs/CMP.csv',
                        help='1st data into the system')

    parser.add_argument('--data1', '-d1',
                        default='data/csvs/SLCA.csv',
                        help='2nd data into the system')

    parser.add_argument('--fromdate', '-f',
                        default='2015-01-01',
                        help='Starting date in YYYY-MM-DD format')

    parser.add_argument('--todate', '-t',
                        default='2019-06-01',
                        help='Starting date in YYYY-MM-DD format')

    parser.add_argument('--period', default=60, type=int,
                        help='Period to apply to the Simple Moving Average')

    parser.add_argument('--cash', default=100, type=int,
                        help='Starting Cash')

    parser.add_argument('--runnext', action='store_true',
                        help='Use next by next instead of runonce')

    parser.add_argument('--nopreload', action='store_true',
                        help='Do not preload the data')

    parser.add_argument('--oldsync', action='store_true',
                        help='Use old data synchronization method')

    parser.add_argument('--commperc', default=0.005, type=float,
                        help='Percentage commission (0.005 is 0.5%%')

    parser.add_argument('--stake', default=1, type=int,
                        help='Stake to apply in each operation')

    parser.add_argument('--plot', '-p', default=True, action='store_true',
                        help='Plot the read data')

    parser.add_argument('--numfigs', '-n', default=1,
                        help='Plot using numfigs figures')

    return parser.parse_args()




def find_history(auto_save=None ):
    # Note finaical history is a pandas dataframe
    stocks_of_interest = None
    failed_symbols = []
    financial_history = {}
    pairs_list = []
    if not os.path.isfile('co_integration_stock_history.pkl'):

        start_date = pd.Timestamp(dt.strptime("2005-06-16", '%Y-%m-%d'))
        end_date = pd.Timestamp(dt.strptime("2019-06-01", '%Y-%m-%d'))

        with open('stocks_of_interest.pkl', 'rb') as handle: 
            print('loading_from_file')
            stocks_of_interest = pickle.load(handle)


        for stock in stocks_of_interest:
            print('Getting ', stock)
            try:
                financial_history[stock] = pdr.get_data_yahoo(stock,
                                                                  start_date,
                                                                  end_date)

                print('Loaded ', len(financial_history[stock]), ' histories. For company ',stock,'.')
            except (ValueError) as e:
                print('ERROR: Failed to get history for', stock)  
                failed_symbols.append(stock)

                time.sleep(1)


        for i, l in enumerate(failed_symbols):
            if i == 0:
                print('Failed to load: ')
            print('\t',l)

        with open('co_integration_stock_history.pkl', 'wb') as handle:
            pickle.dump(financial_history , handle, protocol=pickle.HIGHEST_PROTOCOL)


    else:
        with open('co_integration_stock_history.pkl', 'rb') as handle: 
            print('loading_from_file')
            financial_history = pickle.load(handle)

        with open('stocks_of_interest.pkl', 'rb') as handle: 
            print('loading_from_file')
            stocks_of_interest = pickle.load(handle)   

        with open('stock_pairs_of_interest.pkl', 'rb') as handle: 
            print('loading_from_file')
            pairs_list = pickle.load(handle)   
            print(pairs_list)


    return stocks_of_interest,financial_history, pairs_list


if __name__ == '__main__':
    stocks_of_interest,financial_history, pairs_list = find_history(auto_save=None)
    runstrategy(stocks_of_interest,financial_history, pairs_list)
'''

[ab_trader]

@east1992 said in AttributeError: 'str' object has no attribute '_name' - The error is clear - but for the life of me I can't see why it's happening:

I can attach the code if that would be helpful

Everybody here can read other people minds, sure no script is needed.

[Capn_Hacky]

@ab_trader As you like

[backtrader]

@ab_trader said in AttributeError: 'str' object has no attribute '_name' - The error is clear - but for the life of me I can't see why it's happening:

I can attach the code if that would be helpful

Everybody here can read other people minds, sure no script is needed.

@east1992 said in AttributeError: 'str' object has no attribute '_name' - The error is clear - but for the life of me I can't see why it's happening:

@ab_trader As you like

I think @ab_trader was just trying a prank ...

In any case ... let me wonder how you write such a script, yet miss the top of the forum

For code/output blocks: Use ``` (aka backtick or grave accent) in a single line before and after the block. See: http://commonmark.org/help/

Without the three backticks your code is basically rendered in a form which cannot be properly read.

@east1992 said in AttributeError: 'str' object has no attribute '_name' - The error is clear - but for the life of me I can't see why it's happening:

  File "C:\ProgramData\Anaconda3\envs\finance\lib\site-packages\backtrader\broker.py", line 80, in getcommissioninfo
    if data._name in self.comminfo:
AttributeError: 'str' object has no attribute '_name'

The problem is clear: you have created a position for something which is a string and not a data feed. Hence data (which is a str) has no attribute _name

From what one can read, it is unclear why you need to define a `get_data_by_name', given that a method to retrieve data feeds by name does already exist

• Docs - Strategy - See getdatabyname

But playing with the names is for sure the root of the problem.

[Capn_Hacky]

Thank you for taking the time to answer my question; unfortunately the proposed answer is incorrect. This makes sense if you consider that a custom getter function devised in the manner above shouldn't functionally alter the rest of the code. Specfically I've implimented the suggested change from:

                   p0_data = get_data_by_name(pair[0])
                   p1_data = get_data_by_name(pair[1])

which uses my custom getter function and changed this to

           p0_data = self.getdatabyname(pair[0])
           p1_data = self.getdatabyname(pair[1])

This alteration resulted in an identical error, so indeed is unlikely to have functionally altered anything.