[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Normalization



PureBytes Links

Trading Reference Links


Many variables of Technical Analysis are 
fluctuating in a wide range.
It is more convenient to put them all in the same 
scale, 0...1 or 0...100.
The result of this technique will be a 
normalization of a variable.
Among various ways of normalization we will present 
here the following,
applied over close price.
 
a. Short Term Stochastic
 
/*AFL code for Short Term Normalized 
Close*/
X=CLOSE;
<FONT face=Arial 
size=2>NormalizedX=100*(X-LLV(X,14))/(HHV(X,14)-LLV(X,14));
graph0=NormalizedX;
 
b. Long Term Stochastic
 
/*AFL code for Long Term Normalized 
Close*/
X=CLOSE;
<FONT face=Arial 
size=2>NormalizedX=100*(X-LOWEST(X))/(HIGHEST(X)-LOWEST(X)); 
<FONT face=Arial 
size=2>graph0=NormalizedX;
 
The result is a graph oscillating in the range 
0...100.
Replacing CLOSE in above codes with another 
variable, you will normalize it.
(X=VOLUME, X=MACD(), X=CCI(14), etc.)
Note that some functions of T/A are already 
normalized by definition
(RSI(PERIODS), STOCHK(PERIODS), STOCHD(PERIODS) 
etc)
and do not need further normalization.
The above transformation leads to interesting 
results.
 
Dimitris Tsokakis
 
P. S. If your data extend more than two years, you 
may replace 
LOWEST(X) with LLV(X,500) respectively HIGHEST(X) 
with HHV(X,500).
Support and Resistance levels, formated 3 years 
ago, perhaps are not valid anymore.