Various - page 89

 
mntiwana:

Dearest MLADEN

Thanks for this new style,do you feel and noticed the lower dim color lines are a bit smoother than upper ones.

regards

mntiwana

It entirely depends on the data - I guess that in some trend change process, the loss of a trend is more straight forward then the forming of the new trend

 

This is the " Middle-High-Low Moving Average" as described by Vitali Apirin. It is not a single average but rather a correlation of two averages : one being a regular average and the other is an average of highest high and lowest low. Their crosses are showing the current "trend". Used the default parameters proposed by Vitali Apirin - seems that they are rather OK, except by default he is using SMA while I decided to set EMA as the default average method, but I do recommend some experimenting with the parameters.



Files:
 
mladen:

QQE (any symbols version) with 20 types of prices and 7 types of filters


QQE with the addition of on-chart arrows


 

There was at one stage Laguerre RSI variation indicator. Attaching it to the post (for comparison reasons). But today I have noticed that it can not produce the original Laguerre RSI value. So, I decided to make a new one and to extend it. This is the new version.

It can (when default values are used) produce the same original Laguerre RSI (data level depth set to 2 and iterations set to 1)  and has extra : price smoothing prior to using and price filtering. Price filtering can produce some interesting results. Calculation iteration makes smoother and smoother results as the iteration count rises


For comparison : upper is the original Laguerre RSI value, lower is the same Laguerre rsi with iteration count set to 3 and filter set to 3.


PS: keep in mind that the "variation 2" is posted for comparison purposes only - they are not the same

 
mladen:

There was at one stage Laguerre RSI variation indicator. Attaching it to the post (for comparison reasons). But today I have noticed that it can not produce the original Laguerre RSI value. So, I decided to make a new one and to extend it. This is the new version.

It can (when default values are used) produce the same original Laguerre RSI (data level depth set to 2 and iterations set to 1)  and has extra : price smoothing prior to using and price filtering. Price filtering can produce some interesting results. Calculation iteration makes smoother and smoother results as the iteration count rises


For comparison : upper is the original Laguerre RSI value, lower is the same Laguerre rsi with iteration count set to 3 and filter set to 3.


PS: keep in mind that the "variation 2" is posted for comparison purposes only - they are not the same

Dearest MLADEN

 Thanks for upgrading/extending variation ver to new with smooth and filters,i tried much to get close your setting that is so perfect even with default with iteration count and filter to 3,but not able to create as yours,may be due to there are other factors too, prices,filter period and gamma too that i don't know what setting you used :)

regards


 

The "new" qqe with the addition of the 6 types of rsi (it already has 7 types of filters, alerts, arrows, multi time frame and multi symbol support)


 

MHL that can use averages (the latest set of averages) for calculation (that includes double smoothing too)


Files:
MHL averages.ex4  108 kb
 
mladen:

This is the " Middle-High-Low Moving Average" as described by Vitali Apirin. It is not a single average but rather a correlation of two averages : one being a regular average and the other is an average of highest high and lowest low. Their crosses are showing the current "trend". Used the default parameters proposed by Vitali Apirin - seems that they are rather OK, except by default he is using SMA while I decided to set EMA as the default average method, but I do recommend some experimenting with the parameters.



Middle-High-Low Moving Average made as an oscillator


PS: will use this post for something unrelated. At one forum I have read that "all highest high lowest low based indicators repaint". Quite frankly I do not know if I have read a bigger nonsense till now. Just to remind all : stochastic is based on highest high and lowest low and as far as I know, there is no repainting stochastic (yet :)). So, the fact that this indicator uses highest high and lowest low is not making it a repainter. It does not repaint and the indicator (the regular version or the on chart version) will never repaint

 

This is a sort of regression explorer, ...

What it does is the following : calculates regression of any degree, converts that value into an "end-pointed regression" (ie : average, ie: non-recalculating, non-repainting value) draws that as a kind of average, and for the current bar it draws the regression channel too, including the extrapolated (future) projection of the current channel. So, the "average" itself does not recalculate/repaint (ever) but the channel itself is doing what any regression channel does : it fits itself (recalculates) depending on the current price changes

As a comparison : if you use degree 1, then it is a linear regression and the values in that case have to be exactly the same as linear regression sketcher (if the parameters are the same)


In short : this can be treated as an end-point (non-repainting) regression too (not just as a "regression explorer")


Files:
 

Some purely esthetic changes (the way how display is handled). values are the same