# [NM]Improved Linear Regression Bull and Bear Power v02

12397 lượt xem
Hi guys, I'm back with a little improvement on the Bull and Bear Signal I published just last week thanks to some feedback I received from a couple of users, which is of course highly appreciated.

Here are the changes that have been implemented compared to v01 :
(version 1 is the top indicator, version 2 is the bottom one) in the chart above

• Formula adapted to calculate the signal if no data is available for either bull or bear
• Added the possibility to smoothen the signal using Arnaud Legroux Moving Average (the benefit of this is that it does not add any lag to the signal)

If you have any further ideas on how to improve the indicator or if you are happy with it and want to share your settings or rules of engagement, please feel free to share them below.

Oh, and don't forget to click that like button ! :)
```//@version=2
// this code uses the Linear Regression Bull and Bear Power indicator created by RicardoSantos
// and adds a signal line
// Use : if signal line is changes color, you have your signal, green = buy, red = sell
// Advice : best used with a zero lag indicator like ZeroLagEMA_LB from LazyBear
// if price is above ZLEMA and signal = green => buy, price below ZLEMA and signal = red => sell
// ***** Changelog compared to v01 ******
// Adapted formula to calculate the signal in case there is no information for either bear or bull
// Added the possibility to smoothen the signal (this is done by a simple SMA)
study(title='[RS][NM]Improved Linear Regression Bull and Bear Power v02', shorttitle='BBP_NM_v02', overlay=false)
window = input(title='Lookback Window:', type=integer, defval=10)
smooth = input(title='Smooth ?', type=bool, defval=true)
smap = input(title='Smooth factor', type=integer, defval=5, minval=2, maxval=10)
sigma = input(title='Sigma', type=integer, defval=6)

f_exp_lr(_height, _length)=>
_ret = _height + (_height/_length)

h_value = highest(close, window)
l_value = lowest(close, window)

h_bar = n-highestbars(close, window)
l_bar = n-lowestbars(close, window)

bear = 0-(f_exp_lr(h_value-close, n-h_bar) > 0 ? f_exp_lr(h_value-close, n-h_bar) : 0)
bull = 0+(f_exp_lr(close-l_value, n-l_bar) > 0 ? f_exp_lr(close-l_value, n-l_bar) : 0)
direction = smooth ? alma(bull + bear, smap, 0.9, sigma) : bull*3 + bear*3
dcolor = smooth ? direction[0] > direction[1] ? green : direction[0] < direction[1] ? red : yellow : direction > bull ? green : direction < bear ? red : yellow

plot(title='Bear', series=bear, style=columns, color=maroon, transp=92)
plot(title='Bull', series=bull, style=columns, color=green, transp=92)
plot(title='Direction', series=direction, style=line, linewidth=3, color= dcolor)
plot(0,title='zero line', color=black, linewidth=2)

```

## Bình luận

Hi Nico, @Nico.Muselle. Thanks for this script. I have adapted it by plotting another line for a long term resolution but I have observed that the color on the long term resolution line is not accurate on real time data as compared to historical data. Here is the code. Do you know how I can fix the problem?

//@version=2
// this code uses the Linear Regression Bull and Bear Power indicator created by RicardoSantos
// and adds a signal line
// Use : if signal line is changes color, you have your signal, green = buy, red = sell
// Advice : best used with a zero lag indicator like ZeroLagEMA_LB from LazyBear
// if price is above ZLEMA and signal = green => buy, price below ZLEMA and signal = red => sell
// ***** Changelog compared to v01 ******
// Adapted formula to calculate the signal in case there is no information for either bear or bull
// Added the possibility to smoothen the signal (this is done by a simple SMA)
study(title='Improved Linear Regression Bull and Bear Power v02', shorttitle='Linear Regression Bull/Bear Power', overlay=false)
window = input(title='Lookback Window:', type=integer, defval=10)
smooth = input(title='Smooth ?', type=bool, defval=true)
smap = input(title='Smooth factor', type=integer, defval=5, minval=2, maxval=10)
sigma = input(title='Sigma', type=integer, defval=6)
longTermResolution = input(title='Long term resolution', type=resolution, defval='D')

f_exp_lr(_height, _length)=>
_ret = _height + (_height/_length)

h_value = highest(close, window)
l_value = lowest(close, window)

h_bar = n-highestbars(close, window)
l_bar = n-lowestbars(close, window)

bear = 0-(f_exp_lr(h_value-close, n-h_bar) > 0 ? f_exp_lr(h_value-close, n-h_bar) : 0)
bull = 0+(f_exp_lr(close-l_value, n-l_bar) > 0 ? f_exp_lr(close-l_value, n-l_bar) : 0)
longTermBear = security(tickerid, longTermResolution, bear)
longTermBull = security(tickerid, longTermResolution, bull)

direction = smooth ? alma(bull + bear, smap, 0.9, sigma) : bull*3 + bear*3
colorSeries = smooth ? direction > direction ? 0 : direction < direction ? 1 : 2 : direction > bull ? 0 : direction < bear ? 1 : 2
dcolor = colorSeries == 0 ? lime : colorSeries == 1 ? maroon : yellow

longTermDirection = smooth ? alma(longTermBull + longTermBear, smap, 0.9, sigma) : longTermBull*3 + longTermBear*3
longTermColorSeries = smooth ? longTermDirection > longTermDirection ? 0 : longTermDirection < longTermDirection ? 1 : 2 : longTermDirection > longTermBull ? 0 : longTermDirection < longTermBear ? 1 : 2
longTermColor = longTermColorSeries == 0 ? green : longTermColorSeries == 1 ? red : yellow

divergence = (longTermDirection < 0 and direction > 0 and colorSeries != colorSeries) or
(longTermDirection > 0 and direction < 0 and colorSeries != colorSeries) ? true : false

plot(title='Bear', series=bear, style=columns, color=maroon, transp=92)
plot(title='Bull', series=bull, style=columns, color=lime, transp=92)
plot(title='Direction', series=direction, style=line, linewidth=4, color= dcolor, transp=65)

plot(title='Bear', series=longTermBear, style=columns, color=maroon, transp=92)
plot(title='Bull', series=longTermBull, style=columns, color=green, transp=92)
plot(title='Direction', series=longTermDirection, style=line, linewidth=4, color= longTermColor)

barcolor(divergence ? yellow : na )

plot(0,title='zero line', color=divergence ? blue : orange, linewidth=2)

Phản hồi
Genial Nico. Thanks a lot +10
Phản hồi
AndrewHook
Thanks ! :)
Phản hồi
All I can say is wow. . . I did a little manual back testing on UJ daily using the default settings, changed the chart to Renko , and added LazyBear's ZeroLagEMA you suggested in the code comments. Following your instructions in the code comments results in a very high Win %, Great RR on each trade, and a very simple system.

I do not have a PRO account. So, I am unable to test on any time frame lower than D but I imaging there will be more false signals/losses on lower time frames. However, if further analysis supports the signals, lower time frames may still be quite profitable.

Thank you for you hard work on this indi!
Phản hồi
Great improvement!
Phản hồi
Thanks a lot ! :)
Phản hồi
VERY GOOD
Phản hồi
hi ,

Can i use it for Daily chart as well? Love your work , brother !

thank you so much. Have a nice weekend.
Phản hồi
wow! great work i gona find good indicater
Phản hồi
Have a question on your renko chart here I assume that your setting is 10 pips would that work on the 4 hour chart as well? PS I do like your linear program.
Phản hồi