Seeking advice for recreating an EFI w ATR indicator

Hi Jason,

What exactly do you want to feed into the ATR indicator?

It is not yet clear to me what you are doing here exactly but maybe this could be helpful to you 

https://lean-api-docs.netlify.app/classQuantConnect_1_1Indicators_1_1IndicatorExtensions.html#:~:text=%E2%97%86%C2%A0-,Of%3C%20T,-%3E()

Thank you. I have looked into IndicatorExtensions but am unsure how to use them in this context. Basically, for every point on an EMA, I need to compute the difference from the previous EMA value and then use that to compute the ATR channels from the EMA line.

Could you explain what the ATR channels are? Are they just multiple “normal” ATR indicators?

Hi Jason, Mak,

Just wanting to add a lil something here.

I am using line 74 in the TV indie: atr_ema = math.abs(efi[1] - efi)

I am guessing you are stuck on trying to find a way to access efi[1] and efi. You are saying you have a deque.

So:

• If you append to your deque (append()), the most recent values are on the right, and thus you would do: abs(my_deque[-2] - my_deque[-1])
• If you appendleft to your deque (appendleft()), the most recent values are on the left, and thus you would do: abs(my_deque[1] - my_deque[0])

Of course, assuming the name of your deque variable is “my_deque”.

Now that you know out to get the values for the series passed to RMA on line 75 (atr_out =  RMA(atr_ema, calc_type == "Fixed" ? len : len_out_slow) ), you just need to create an RMA indicator (Wilder's MA) and use its Update() method to push those values to it. Once it is ready, you're ready.

In fact, I first thought RMA was not already in QC's indies because I searched for RMA. But it looks like it is WilderMovingAverage. Create it using the long name “WilderMovingAverage” so you can update it yourself instead of LEAN updating it on each bar (you do not want LEAN to update it with bar data but you updating it with distance data (abs(old_value - recent_value) is distance between values).

Fred

Fred, thank you. You understand what I mean. I'll unpack your suggestion and try it out this evening.

Mak, ATR channels are a factor of 1, 2, and 3 times the EMA delta away from the EMA line. The TV code that I linked to in the original post probably explains it better.

A very similar picture.

Used self.History and talib functions. 

Let me know if you will see it...

Ya, I'm not sure how to get past the block I'm experiencing. I've attached the latest backtest. Thinking it's foobared and need to start fresh.

Jason Annable,

As I mentioned above, it is fairly easy to calculate the Elder Force Index using the self.History  and talib  functions.

What is related to the TV code of ATR channels it has nothing to do with the Average True Range Indicator.

So I decided to recreate these channels using talib. ATR () and talib.SMA (volumes).

The picture is very similar to the one on TV.

Happy Holidays.