SAP Knowledge Base Article - Public

1864206 - Using group sorting and top percentage gives incorrect results in Crystal Reports

Symptom

  • Unable to change the percentage value in the Group Sort Expert for Top Percentage, as required
  • Incorrect results are displayed
  • The sorting is changed incorrectly

Environment

  • Crystal Reports 2011 Support Package 4 Patch 7
  • Crystal Reports 2011 Support Package 5

Reproducing the Issue

  1. Insert a group based on Material for example
  2. Go to Group sort expert > Select the "Top percentage" sort > Set the percentage value as 26.93 for example (observe that the first group with percentage more than 26.93 appears and all others are grouped together)
  3. Again go to Group sort expert > Set the percentage value as 26.94 this time

Observations:

Now the first 2 groups are displayed and all the other records are placed in the Others Group. But as only the first group is above 26.94% the sorting must remain the same.


Cause

This has been identified as a potential bug and escalated to Product Group under reference number ADAPT01694417.

Resolution

Upon investigating the issue Product Group has come to the conclusion that this issue happens by design due to the logic of the current algorithm.  Its been debated whether this logic needs to be reconsidered taken to overcome this issue but after weighing the pros and cons of the change, they have come to the conclusion that the introduction of a new algorithm would be too risky overall.

 

Below are the details of each benefits vs the shortcomings: 

Current algorithm:

  1.  Benefit

1>     It exists since a long time, and at present is accepted by a majority of customers.

2>     The algorithm can handle all the cases:

  • For VGroupn > 0  and ( VGroup1 + VGroup2 + ……….+VGroupn) >0    (n= 1, 2, 3,…..n), it’s the same as the new algorithm
  • For VGroupn < 0  and ( VGroup1 + VGroup2 + ……….+VGroupn) =0    (n= 1, 2, 3,…..n), it can handle
  1. Shortcoming

1>     For VGroupn < 0  and ( VGroup1 + VGroup2 + ……….+VGroupn) <0    (n= 1, 2, 3,…..n), that is, it’s not precise

 

New algorithm

  1. 1.       Benefit

1>     For VGroupn < 0  and ( VGroup1 + VGroup2 + ……….+VGroupn) <0    (n= 1, 2, 3,…..n), it’s precise.

  1. 2.       Shortcomings

1>     It’s a new algorithm, most customers use accept it and have been using it since long.  Changing the logic might introduce new problems and hence it a risk factor;

2>     As it’s a new algorithm, maybe for some cases, it might still not satisty a condition again introducing a risk factor;

3>     For  case VGroupn < 0  and ( VGroup1 + VGroup2 + ……….+VGroupn) =0    (n= 1, 2, 3,…..n), it cannot handle and some customers might not accept this;

4>     If the new algorithm is put in place, customers would be allowed to input any value for Percentage(at present, only (0~100) can be selected), and for some customers this might be consfusing.

 

Conclusion

A new algorithm might introduce many issues and hence the old algorithm will be retained.

Keywords

KBA , BI-RA-CR , Crystal Reports designer or Business View Manager , Bug Filed

Product

Crystal Reports 2008 V1 ; SAP Crystal Reports 2011