Characteristic Weight/Defect Class - no decimals

[Former Member]

Anyone know if there is an enhancement that would allow for decimals to be used? I have an extraordinary request that calls for certain qualitative scores (characteristic weights) to be decimals.

[Former Member · Accepted solution]

Char weitage can not be in decimal places actually.

What is exact requirement.You may use some user exit in case they will satisfy your querry like

QEEM0002 |User exit add. functions after reading inspection features |

| QEEM0003 |User exit: add. functions after valuating insp. characs

[Former Member · Accepted solution]

Thanks SujitSND, my requirement is for an inspection that has about 100 characteristics. Each one has a weight (about 9-10 values, some 1, some 2, some 3.5, etc). These values only appear when the qualitative inspection is "No", which then should be multiplied by the defect class value. If the inspection "Yes", then a valuation of 0 should appear. These inspections need to be qualitative because no measured value is actually taking place. Would you happen to have a way to address this? Thanks again!

[former_member42743 · Accepted solution]

Is this for calulating an overall "score" for the material/inspection lot?

You can define characteristic weightings in config and then use them in your defect types so that each defect type can have a different weighting. SAP can then calculate the overall Q score for the lot on this basis giving you weighted score for the lot.

Is that the type of thing you are looking for?

FF

[Former Member · Accepted solution]

Yes, except that I want to be able to use decimal numbers (to replicate the legacy method that is currently being performed in excel). <p>Also, which score calculation do you recommend? The ones that come standard give meen average scores (or others) I believe. Is there one that gives a cummulative total? Or, will a developer have to use the existing functional module to achive this? Thanks for your help btw.

Edited by: PMPSMM on Apr 5, 2010 1:50 PM

[former_member42743 · Accepted solution]

I'm not sure I have a good answer for you on this. You can create a qualitative mic with values such as PASS and a number of different defect types, each of which would each have their own value. That is easy. You could store the weighting value (1, 2.0, 3.5 etc) in one of the available info fileds for the characteristic. Then they would select either pass or a defect code.

To get an overall score, I think you will need a custom formula code. Instead of C00010 you would create a Z1xxxx formula parameter and pass in the first characteristic number. Z10100 for instance. The FM behind Z1 would then check all the characteristics from 0100 to say 0300, (or some range you select). For any it finds with a reject valuation, it mutliples the defect weighting by the value held in the inspection characteristic info field. It adds all these up and passes this back to the calculated characteristic.

I think that is how I'd try my initial approach to this.

FF

[Former Member · Accepted solution]

One thing I thought was true, was that the characteristic weight gets multiplied by the defect class weight, no? Then (according to my initial thought) we could develop a functional module based on KQD_COPY_REFERENCE_FORM to total up the cummulative score. Would that work? (without decimals)

[former_member42743 · Accepted solution]

It might work. I can't say as I never have undertaken something like that. But your Q-score can be calculated by characteristic weighting. But I'm not sure about the defect weight. But since your free to copy the FM used and modify it, you can probably get it to calculate what ever you want. That would give you a Q-score for the lot, but then I think you want that visibility on a batch record right? I'm not sure how you would do that except with maybe a follow-up action on the UD code that would copy the Q-score into the batch record onto a characteristic.

FF

[Former Member · Accepted solution]

This is what the SAP help states in SPRO:

"Weighting Factor for Evaluations

Determines the proportion the individual valuation criteria contribute to the calculation of the quality score.

Example

Other criteria, in addition to the defect, can be used to calculate the quality score. For example, the characteristic weight provides a factor with which the defect class score is multiplied:

Defect class Characteristic

score weight

Characteristic 1 80 X 5 = 400

Characteristic 2 50 X 2 = 100

Characteristic 3 90 X 3 = 270

__ ___

Quality score 10 770 = 77"

So would this not accomplish the requirement (minus the decimal numbers)?

[former_member42743 · Accepted solution]

It looks and sounds like it would. But since I haven't personally tried that I have to leave it up to you to test and check.

What I see your next problem being is that if you had no defects in your legacy system, the value should be 0. So in practice, the higer the score, the worse the product.

In SAP, the Q-score is the opposite. 100 is best, 1 is worst. So your 77 really should become a 23. I'm not sure you'l be able to give a zero value to your score when no defects are recorded. SAP might already do the calculation for you in their module, I'm not sure how that works.

It looks like from the FM KQD_FEATURE_CLASS_WEIGHTED that the worse the defect the higher the number would be for that characteristic. But since SAP defines 100 as the best score, that wouldn't work. The module states that if no defect is recorded the charateristic gets its maximum value. So I'm guessing that with the provided SAP module you shoud find that if your example is used, the Q-score would be 23. If no defects are recorded the score should be 100.

Keep us informed about your progess. I'm very interested to hear how this works.

FF

[Former Member · Accepted solution]

Yes, the higher the score, the worse. Actually, they don't use a 100% scoring system. The highest possible score is approx 330. So now this is bad news for me if the Quality Score is not allowed to surpase 100. IMO, I think the way SAP has it setup is better. From when we are in school, we are used to 100 being the best, not 0. I am certainly lobbying for these semantics, and would rather have the scores setup this way. I'll keep you posted on what happens, I certainly appreciate the help!