Hi Thomas,

I have used this in Pipeline:

close_weekly = USEquityPricing.close.latest.downsample('week_start')  

Also in def(close):

close_weekly = lambda close: close.downsample('week_start')  

Quantopian has more information on Filter and Classifier.

Hope this helps.

Hi Karl,

Many thanks! Seems the Q has updated and expended their API Reference.

I will have a try.

Now, I haven't created a weekly CustomerFactor. I do somewhat different. It seems easier but I am not sure if this is correct.

What I did: I haven't applied the downsample onto the close, but on the output indicator value.

I can see your approach, Thomas and I can think of another way is to do lambda close: close.downsample('week_start') on USEquityPricing.close inside your MyTalibKAMA_Weekly(CustomFactor) and cross check both to see if the values are as expected for your purpose.

Hope this helps.

I've changed the MyTalibKAMA_Weekly as follow. There is any error but the outputs are not correct.

class MyTalibKAMA_Weekly(CustomFactor):

inputs = (USEquityPricing.close,)

inputs = (lambda close: close.downsample('week_start'),)  
params = {'kama_len' : 10,'day_offset':0,}  
window_length = 100 # 10 times of kama_len  
window_safe = True  
def compute(self, today, assets, out, close, kama_len, day_offset):  
    # TODO: figure out how this can be coded without a loop  
    kama = []  
    for col in close.T:  
        try:  
            kama_tmp = talib.KAMA(col, timeperiod=kama_len)  
            kama.append(kama_tmp[-1-day_offset])  
        except:  
            kama.append(np.nan)  
    out[:] = kama  

Hey! Glad to see somebody with the same isssue. Could you give me a helping hand with this custom factor? Seems like your approach doesn't work.

class GetMacDHistSlope(CustomFactor):  
    inputs = [USEP.close] #USEP = USEquityPricing  
    #inputs = [USEP.close.latest.downsample('week_start')]  
    inputs = [lambda close: close.downsample('week_start')]  
    window_length = 400  
    def compute(self, today, assets, out, close):  
        hists = []  
        for stock_close in close.T:  
            try:  
                #Compute only end of week equity prices!!  
                macd, signal, hist = talib.MACD(stock_close, fastperiod=12,  
                                                slowperiod=26, signalperiod=9)  
                if (hist[-3] > hist[-2]) and (hist[-2] > hist[-1]):  
                    trend = -1  
                elif (hist[-3] < hist[-2]) and (hist[-2] < hist[-1]):  
                    trend = 1  
                else:  
                    trend = 0  
                hists.append(trend)  
            except:  
                hists.append(np.nan)  
        out[:] = hists  

It gives me as a result

AttributeError: 'NoneType' object has no attribute 'format'