The error is saying the 'security' object doesn't have a method or attribute named USEquityPricing. The exact error is in this line:

if security.USEquityPricing.price.latest < sma_50:

However, there is another issue with the way 'SimpleMovingAverage' is being used. That needs to be put into the pipeline definition. Take a look at the tutorials if you haven't already (https://www.quantopian.com/tutorials/getting-started). Lessons 5-7 demonstrate how to create the pipeline definition, get the results in a dataframe, and then use those results in logic. I've also attached a simple algo with these steps which may get you moving forward.

Hope that helps. Good luck.

Dan,

Thanks for the help.

I have moved the calculation of moving average down to pipeline. Now it's a matter of checking the condition on securities that are already in the portfolio.

I'm getting this error now:

TypeError: 'zipline.assets.assets.Equity' object has no attribute 'getitem_'
USER ALGORITHM:26, in my_rebalance
if security['latest price'] < security['50D MA']:

Are you trying to run an algo within a notebook or did you just copy your algo code into a notebook for demonstration? Algos don't run in notebooks and it would be helpful if you attached the actual algo. That said, a couple of things.

Looking at this statement

    for security in context.portfolio.positions:  
        # check to see if our longs have broken down. otherwise, do nothing.  
        if security['latest price'] < security['50D MA']:  
            order_target_percent(security,0)  
        else:  
            continue

'security' is an equity object. It only has a limited number of attributes (see https://www.quantopian.com/help#api-sidinfo). The fields you're trying to reference ('latest price' and '50D MA') are columns in the pipeline output (ie context.output) which is a pandas dataframe. The best way to fetch single values from a dataframe is to use the .at method (https://pandas.pydata.org/pandas-docs/version/0.18/generated/pandas.DataFrame.at.html). So maybe the code could be something like this:

    for stock in context.portfolio.positions:  
        yesterday_close = context.output.at[stock, 'latest price']  
        sma_50 = context.output.at[stock, '50D MA']  
        if yesterday_close < sma_50 and  data.can_trade(stock):  
                order_target_percent(stock, 0)

That fetches 'latest price' and '50D MA' for that specific security from the context.output dataframe. It then compares those values. BTW it's always a good idea to check if 'data.can_trade' in case a security became delisted. Doesn't happen often but the algo will error if it does.

While the above should work, some may consider it not very 'pythonic'. Generally, python (and especially pandas) is better at doing things across entire groups of data in a parallel manner and then working with those groups. Additionally, it's often very advantageous to NOT use spaces in dataframe column names. Best to use an underscore instead. So this would be another way to do the same thing:

    # Select all the stocks meeting the query criteria then turn that into a list  
    close_these = context.output.query('latest_price < 50D_MA').index.tolist()

    for stock in context.portfolio.positions:  
        if stock in close_these and data.can_trade(stock):  
            order_target_percent(stock, 0)

Part of the benefit of this approach is by using the query method the logic can be as involved and complex as one wishes and still remain very readable. See https://pandas.pydata.org/pandas-docs/version/0.18/generated/pandas.DataFrame.query.html and https://pandas.pydata.org/pandas-docs/version/0.18/indexing.html#indexing-query for some good examples.

Hope that get's the error resolved.

Dan,

Thanks so much. I was able to get it working using this:
close_these = context.output.query('latest_price < 50D_MA').index.tolist()

I just put it in a notebook since it wouldn't run a backtest because of the errors.

I've attached one now. More work to be done!

Thanks,
Andrew

Andrew, glad to help. Looks good.