Problem in Ticket Index Digits Count Filter

[Red Devil]

Joe,

I've attached screenshots of different settings and results in this filter. Either I've misunderstood the filter or there appears to be a small problem with the results achieved.

a). I've set it to filter for 3 x digit 6 leaving 287887 tickets
b). I've set it to filter for 3 x digit 9 leaving 284642 tickets
c). I've then set it to filter either of the above by selecting count min=1 max=1 leaving 570137 tickets

I would have thought that c would be the sum of a & b; 572,529 not 570,137.

Is it me or an error with the filter?

Rgds

Red Devil

[Joe]

[quote=Red Devil]
Joe,

I've attached screenshots of different settings and results in this filter. Either I've misunderstood the filter or there appears to be a small problem with the results achieved.

a). I've set it to filter for 3 x digit 6 leaving 287887 tickets
b). I've set it to filter for 3 x digit 9 leaving 284642 tickets
c). I've then set it to filter either of the above by selecting count min=1 max=1 leaving 570137 tickets

I would have thought that c would be the sum of a & b; 572,529 not 570,137.

Is it me or an error with the filter?

Rgds

Red Devil
[/quote]

will check it Red, thanks,

[Joe]

Red,

Please try this jar file, I hope I fixed the problem, Now you can use this in compound filter,

Please let me know what you expecting,

Thanks

[Jerzy]

Red Devil and Joe,

After seeing the screnshots I realise that in each of 3 examples of filtering you filtered the full package of tickets. Three full packages of tickets were filtered independently, and the result of filtering in each case was correct. Unfortunatelly, adding the results from points a) and b) and expecting that the sum would be exactly as in c) is methodologically incorrect because filtrations in the case a) and b) were independent from each other.

Try to filter first like in the point a), and then use the settings like in the point b) to filter further the prefiltered package (not the full package). Only this way you can achieve the same result as in the point c). Ranges of two settings overlap each other to some extent and the resulting effect is smaller than expected (or smaller than it would be when filtering independent packages).

Best Regards

Jerzy

[Joe]

Thanks Jerzy, I didn't see any bugs in calculations there, but that jar file will not work in compound filter as I didn't do anything for compound filter,

Here is next version of this filter, i fixed the bugs in compound filter, slight changes in previous one,

Thanks

[Red Devil]

[quote=Jerzy]
Red Devil and Joe,

After seeing the screnshots I realise that in each of 3 examples of filtering you filtered the full package of tickets. Three full packages of tickets were filtered independently, and the result of filtering in each case was correct. Unfortunatelly, adding the results from points a) and b) and expecting that the sum would be exactly as in c) is methodologically incorrect because filtrations in the case a) and b) were independent from each other.

Try to filter first like in the point a), and then use the settings like in the point b) to filter further the prefiltered package (not the full package). Only this way you can achieve the same result as in the point c). Ranges of two settings overlap each other to some extent and the resulting effect is smaller than expected (or smaller than it would be when filtering independent packages).

Best Regards

Jerzy
[/quote]

Jerzy,

Yes you are correct, in each of the single filters there will be cases when there are 3 of each ( 6&9). Obviously did not have my thinking head on. :vogel:

Thanks