Request extension to Highest/Lowest operator

[tdnl46w]

3. - Statistical Filter.
Add new operator for highest and lowest per table row instance; Highest Rows/Lowest Rows(or something similar). For example. If you sort an analyzer table column, the 10 highest occurrences can each have a value of "5.00". But with this operator they can still be filtered as 10 separate instances using the table's sorting order. This would be very useful for strategies that require separate properties(to break up numerical ties etc.) or statistical combinations.

...

I can imagine an extension to the current highest/lowest operators to select x lowest/highest UNIQUE values. Is that what you meant?

Yes, as it would make it possible to correctly filter/analyze some combinations or strategies that require unique values.

[dreamweaver]

Hi Stan,

I just wanted to add to tdnl46w's request for this feature inclusion. I am not sure if tdnl46w's feature is along the same line as what I have interpreted from his request but
what I would like to see is when you specify a max (highest) and min (lowest) pass rates within most of EL5's Filters what I would like to see is to have the ability refine or de-select values
that may exist within this string value range.

To put this into very simplistic terms let's say that in one of EL5's Filter you may set a max pass rate value of 20 and a min pass rate of 9.

So the values in this range are extrapolated as follows:

20,19,18,17,16,15,14,13,12,11,10,9

What I would like to have available is the ability to del-select or take out one or more of the values in the above range

For example: you may want to exclude the following values: 19, 16, 13, 11 and 10

Therefore you end up with a modified max and min pass range and the Filter will process against those values only

i.e. 20,18,17,15,14,12 and 9

I would be interested to see this implemented in most of the EL5's Filters and would be more than welcome to discuss
on what Filters could have this feature available and what it would look like.

The benefits of this enables one to reduce the package considerably and possibly eliminate "noise" within the max and pass rate range,
From this example you are not filtering on 12 values but a reduced range of 7.

One could say why not create single lines in a complex filter for each of the 7 values !!! I believe this to be inefficient and
would lead to creating bulky and bloated filters. Why not have this available and achievable in a single iteration.

There are many possibilities for this approach and can also be used more so as well under the Winning Numbers Analyzing Feature

[stan]

So you're suggesting that if a filter has 'min to pass & max to pass' option then you want to remove additional values from that range?

[dreamweaver]

Hi Stan,


Please look at my response in a new thread I have started under the Topic Heading

"Max and Min Pass Rate Enhancements"

[stan]

I can imagine an extension to the current highest/lowest operators to select x lowest/highest UNIQUE values. Is that what you meant?

I've started working on this and after reading the source code again after some time I've realized that it is how the current lowest/highest operators work.

[tdnl46w]

I've started working on this and after reading the source code again after some time I've realized that it is how the current lowest/highest operators work.

Thanks for working on this request. My operator idea was basically just a highest/lowest by table sorting order instead of rank. So it's probably better to leave it as it is.