Hi Chuck,

I don't have time now to tinker with it, but perhaps you could try commenting out sections of code until you can get it to run. Then you can run a full backtest and post the code here (with the "Add Backtest" button). It'll make it easier to debug, especially any problem with nesting of if statements.

Also, which line is causing an error? Please copy it here, rather than referring to the line number.

Grant

Hi Chuck,

Not entirely sure how your algorithm is supposed to work, but on this section:

    if context.first_day or context.new_day:  
        context.day_counter = context.day_counter + 1  
        price = data[context.nflx].price  
     if price + .10 <= data[context.nflx].price <= price + .20:  
        order(context.nflx,100)  
        context.bought = True  

If "context.first_day or context.new_day" is false, then "price" is never initialized, so the next statement ("if price + .10...") will be checking an uninitialized variable. If the "price = data[context.nflx].price..." assignment happens outside of the if block, then the error seems to go away.

Also, this section:
if context.bought = True if data[context.nflx].price <= price - .20: order(context.nflx,-100)

"if context.bought = True" is kind of floating in the ether - either it needs a colon (and a statement to execute), or you should combine it with the following if statement using "and".

Hope that helps!

Thanks John and Grant,

I noticed both these errors as I was commenting out the errors.

I am going to give it another go from scratch line by line.

How would I test for my openp variable?

Seem to be getting an error on the line where I am comparing that value to price before openp is assigned a value.

Try this - I moved "openp" from a local variable (which was getting lost on each successive call to handle_data) to stuffing it into the more persistent "context" object. This preserves it for every iteration.

def handle_data(context, data):

    event_datetime = data[context.stocks[0]].datetime  
    event_day = data[context.stocks[0]].datetime.day  
    # Called only on the very first run.  
    if context.initialize:  
        context.event_day = event_day  
        context.initialize = False  
        context.first_day = True  
    # True at the turn of each new trading day.  
    if event_day != context.event_day:  
        context.new_day = True  
    else:  
        context.new_day = False  
    # If this is the very first event of the day,  
    # mark the open price  
    if context.first_day or context.new_day:  
        context.day_counter = context.day_counter + 1  
        print event_datetime  
        context.openp = data[context.stocks[0]].price  
        print context.openp  
        log.info("New day, open price:  {p} ".format(p=context.openp))  
    else:  
        price = data[context.stocks[0]].price  
        log.info("Current time {t}, price:  {p} ".format(t=event_datetime, p=price))  
        if price > context.openp + .10:  
            print "buy"  
    context.first_day = False  
    context.event_day = event_day  

Thanks John.

That put me on the right track! Now to add more conditions to this algorithm!