consider this

result = indicator1*indcator2*indicator3

If result > entrylevel and seondaryFilter=true then buy...

Indicators will range from -100 to 100 as they are normalized in this range.


Exact mode.

If EntryLevel exact is used, you could use entry level say 0 to 300 step 10 if we have the format below.

Weights 0 to 2 step 0.25

Result=indicator1*weight1+indicator2*weight2+indicator3*weight3

The highest possible value of result would be 100*2+100*2+100*2  = 300

If result{-300 to 300}  > entrylevel {say entry Params 0 to 100 step 10} and seondaryFilter=true then buy...

So this will give a mild effect of making indicator results closer to zero not trade. This will give fewer trades

and possibly some immunity to noisy signals.




Power mode.

consider another example

result = indicator1*weight1*indcator2*weight2*indicator3*weight3

If result > entrylevel and seondaryFilter=true then buy...


The highest possible value of result would be 100*2*100*2*100*2= 8,000,000

So entrylevel of 0 to 300 would be way to low.

However entrylevel of 0 to say 1,000,000 step 10,000 is a bit coarse in its steps too.


This is why we have the power function. Power 4 is used in this example.

Entryparamaters 0 to 10 step 1 would give the values of

0 1 16, 81,  256,  625,  1296,  2401,  4096,  6561,  10000

This is considered more appropriate when we have a wide possible range of numbers in the result variable.


SignPower mode.

Signp(x) gives -1 for a number < 0, 0 for 0, and 1 for a number >0.

If we just used power mode, and used -10 to 10 our entry parameters would be

10000, 6561, 4096, 2401, 1296, 625, 256, 81, 16,1 0 1 16, 81,  256,  625,  1296,  2401,  4096,  6561,  10000


Sign power will in effect do the following when enty Parms are -10 to 10 step 1, entrylevel value ^4

Entrylevel=signp(parameters)*EntryParameters^entryLevelValue.

This would give entry level of

-10000, -6561, -4096, -2401, -1296, -625, -256, -81, -16,-1 0 1 16, 81,  256,  625,  1296,  2401,  4096,  6561,  10000