Br_Squeeze for ninja?

[tazmaan]

Hi everyone. I am very interested in trying out br_squeeze, but I can only locate it for tradestation. Does anyone know where I can find this for ninja, or know of someone that can convert it for me? Thanks!

[NinjaTrader_JoshP]

Hi tazmaan,

You can search the File Sharing indicator section. If you can't find it there then someone has not released it.

You may also try one of the 3rd party NinjaScript Consultants here: http://www.ninjatrader.com/webnew/pa...injaScript.htm

[tazmaan]

Thanks Josh,
I tried to search but will try a few more times I guess. If you happen to have a minute or so and can help me search to make sure I don't miss it would be great.
If noone responds about having it or knowing where it is, i'll give one of the consultants a try.

[anachronist]

Can you post the Tradestation code?

[tazmaan]

Here it is. Please let me know if you can get anywhere with converting it to NT?
Thanks!

Can you post the Tradestation code?

[anachronist]

No, I mean, post the source code, in plain text. Or, point to a description of the general algorithm.

What you posted was an .eld file. It can't be read in any text editor, and TradeStation 2000i (what I have) won't read it. I think .eld is for TS8, not TS2000i.

-Alex

[tazmaan]

Sorry, that's all I have. Is why I was hoping someone knew more about it and if there was one for NT. If I knew how it exactly worked and what algorithm it used I'd be in alot better position with it

No, I mean, post the source code, in plain text. Or, point to a description of the general algorithm.

What you posted was an .eld file. It can't be read in any text editor, and TradeStation 2000i (what I have) won't read it. I think .eld is for TS8, not TS2000i.

-Alex

[mrlogik]

tazmaan.

The code you posted is locked anyway; TS will not allow me to open it.

[jdp2012]

Is this the BB-Squeeze that is being referred to?

Is this the bb-squeeze indicator that is being referred to in this thread. I see it is being called the br_squeeze of which I am not familiiar. J. Carter and H. Senters use a bb-squeeze that they obtained freely from the forums on TradeStation.

[tazmaan]

No, from what I am told br_squeeze is a highly modified version of carter's that actually is useful.

[jdp2012]

Thanks

Tazmaan, thanks for the reply. I am not familiar with the br version. I have TS code for the bb version. But, since the br version actually works, pls let me know if you have or someone has it to share.

[anachronist]

Okay. After researching these 'squeeze' indicators, I have concluded the following, which should be sufficient for anyone to reproduce and improve such an indicator.

1. They all compare two channels. One channel measures trading range over the past N bars (such as Keltner channel), and the other channel measures volatility over the past N bars (such as standard deviation).

2. After picking the channel algorithms, they plot pretty bar charts to indicate when one channel is inside or outside of the other, and whether the difference between them is increasing or decreasing.

That's it.

The generic squeeze indicator uses a Keltner channel and a standard deviation channel centered around a moving average. Actually you don't even need the moving average; all you need are the channel widths.

"Improvements" generally focus on smoothing the channels or making them generate earlier triggers, or refining the calculations.

For example, one could modify the Keltner channel to use average true range, rather than just bar range. You could also substitute a regression channel for standard deviation. In either case, a non-lagging smoother like Hull Moving Average could be applied to it to make it better behaved.

Another improvement is to avoid using mathematical constructs that emit two signals. Standard deviation is one of these. When a large value enters the lookback interval it rightly affects the standard deviation, but it fires again when this large value leaves the interval because all values are equally weighted. In other words, it signals twice, and the second one is false. To fix this, you might use an exponential standard deviation and exponential averages. Or use indicators based on weighted averages.

And for all that, the creators lock down the code and ask for payment. Nothing wrong with that -- they should be compensated for all their time in research and experimentation. However, once you have the framework of such an indicator written, you should do your own experiments to find something that works to your satisfaction. In the end you'll have something that you fully understand, not some canned black box.

-Alex

[PrTester]

Okay. After researching these 'squeeze' indicators, I have concluded the following, which should be sufficient for anyone to reproduce and improve such an indicator.

1. They all compare two channels. One channel measures trading range over the past N bars (such as Keltner channel), and the other channel measures volatility over the past N bars (such as standard deviation).

2. After picking the channel algorithms, they plot pretty bar charts to indicate when one channel is inside or outside of the other, and whether the difference between them is increasing or decreasing.

That's it.

The generic squeeze indicator uses a Keltner channel and a standard deviation channel centered around a moving average. Actually you don't even need the moving average; all you need are the channel widths.

"Improvements" generally focus on smoothing the channels or making them generate earlier triggers, or refining the calculations.

For example, one could modify the Keltner channel to use average true range, rather than just bar range. You could also substitute a regression channel for standard deviation. In either case, a non-lagging smoother like Hull Moving Average could be applied to it to make it better behaved.

Another improvement is to avoid using mathematical constructs that emit two signals. Standard deviation is one of these. When a large value enters the lookback interval it rightly affects the standard deviation, but it fires again when this large value leaves the interval because all values are equally weighted. In other words, it signals twice, and the second one is false. To fix this, you might use an exponential standard deviation and exponential averages. Or use indicators based on weighted averages.

And for all that, the creators lock down the code and ask for payment. Nothing wrong with that -- they should be compensated for all their time in research and experimentation. However, once you have the framework of such an indicator written, you should do your own experiments to find something that works to your satisfaction. In the end you'll have something that you fully understand, not some canned black box.

-Alex

Well, 2 user Post their version in an earlier time, just revise them and add your own taste to them.