Mickey Mouse Feeder Question

[Steve Vincent]

Simplist of rules yet i can't get it to work.

I have 2 values in a cube, total employees and total attrition. All i want is the percentage of attrition, yet the feeder doesn't work as when i suppress zeroes. Must be missing something dozy but i can't see what;

Code: Select all

SKIPCHECK;

['attrition percentage'] = ['Attrition Count'] \ ['Employee Count'];

FEEDERS;
['Attrition Count'] => ['attrition percentage'];

Its a very sparse cube, hence why i'm using skip check. Anyone see why it wouldn't work? Cheers

[Steve Rowe]

Is attrition count rule based and not fed?
Is it an N or a C?

[Martin Ryan]

Any chance of killing the cube and rebuilding it? Looks perfectly fine to me, so perhaps TM1's got its knickers in a knot somehow.

Martin

[nhavis]

The feeder is fine:

Are you looking at the correct cube?

Did you save the rule file?

Why doesn't the feeder work? - Are there non-zero values that are suppressed when they shouldn't be?
All cells with zero in them will be suppressed whether or not they are fed.

[Steve Vincent]

Cheers Guys, the help is much appreciated

• attrition count is loaded via TI, its not rule derived and not a consolidation
• the cube is brand new and i have tried unloading / reloading it but it still misbehaves
• yes i am looking at the right cube (its been known on occasions for me to not be tho!) and yes the rules have been saved.

Are there non-zero values that are suppressed when they shouldn't be?

Yes, thats exactly what the problem is. The numbers are correct, they just get suppressed and on a very sparse cube that is a real pain

[Steve Rowe]

Hmm, very odd

Is that the entire rule sheet?

Maybe try writing the rule like this

['attrition percentage'] =N: ['Attrition Count'] \ ['Employee Count'];

['attrition percentage'] =C: ['Attrition Count'] \ ['Employee Count'];

Double check for typos in the rule or in the naming of the elements in the dimension...
That's me out for things to check, I'd be looking for user error if something as basic as this is going wrong rather than some weird bug....

Cheers

[Martin Ryan]

Have you got the allowseparatenandc rules config parameter turned on or off? Whichever way it is, try doing it the other way.

Martin

[paulsimon]

Steve

You mentioned that ['Attrition Count'] is loaded by TI but what about ['Employee Count'] ? Could this be calculated? Not that this would matter so long as it is not giving a zero result.

Can you set up a view where you can see non-zero figures in both of these, and a value in ['attrition percentage'], when the cube is not zero suppressed?

Does this value only disappear when the view is zero suppressed?

Are these the only rules, ie no possibility of any other rule above interferring?

What version are you running on?

Is this on a server or local server? I recently found problems with two N and C rules, which disappeared when I moved from a local server to a server. Having said that you shouldn't need N and C rules.

Is there possibly an Alias that is causing problems? Even in 9.1 if an Alias is rule derived you can get duplicate aliases or Aliases that are duplicates of other element names, both of which can cause odd problems.

Are you editing the rules via the Rule Editor or Rule Worksheet?

If it's the Rule Worksheet try using the rule editor. Just add a space to make it save.

If using the Rule Editor, try a select all then cut and paste in to Notepad then select all and paste back again. That can help to get rid of any lurking control characters.

Try deleting the rules, and re-entering using the rules wizard - that has been known to cure odd bugs in the past.

Regards

Paul Simon

[Steve Vincent]

Woah! i think the replies to this could turn out to be a FAQ on what to look for if you have issues with rules

Turns out to be none of the above! As i started to dig in to some detailed numbers in the cube i found that the rule was not calculating the percentage lower down, then it suddenly dawned on me why the feeder wasn't working. The Attrition Count is loaded by TI to multiple elements in a particular dimension that details the type of attrition (retirement, redundancy, resignation etc). The employee count (ie. all the people not just attrition) is loaded to a single element in the same dimension (Total Actuals).

So if i looked at the top level of that dim my numbers were fine but got suppressed. As i drilled down i got no percentages at all because the rule was looking for employee count on an element that it would never exist at. All i had to do was change the rule to read;

Code: Select all

SKIPCHECK;

['attrition percentage'] = ['Attrition Count'] \ DB('Attrition','Total Actuals',!Organisation Breakdown,!Resource Breakdown,!Nomination,!Site,!Resource Source,!MonthYear,[b]'Employee Count'[/b]);

FEEDERS;

['Attrition Count'] => ['attrition percentage'];

That fixed the rule lower down and as a result the feeder did its job as i had intended. Its another one of those instances where its easy to miss the fact that feeders only work on the nodal level elements unless you specifically tell them to use a consolidation. Thanks a lot for all the ideas tho, its certainly helped me to get there in the end

[Steve Vincent]

Answers to each of those questions as it may help others in the future;

Steve

You mentioned that ['Attrition Count'] is loaded by TI but what about ['Employee Count'] ? Could this be calculated? Not that this would matter so long as it is not giving a zero result.

No, this is also loaded by TI.

Can you set up a view where you can see non-zero figures in both of these, and a value in ['attrition percentage'], when the cube is not zero suppressed?

Yes.

Does this value only disappear when the view is zero suppressed?

Yes.

Are these the only rules, ie no possibility of any other rule above interferring?

Yes.

What version are you running on?

See my footer and the system details on the right --->

Is this on a server or local server? I recently found problems with two N and C rules, which disappeared when I moved from a local server to a server. Having said that you shouldn't need N and C rules.

Have come across this problem myself in the past, this is running as a server not local.

Is there possibly an Alias that is causing problems? Even in 9.1 if an Alias is rule derived you can get duplicate aliases or Aliases that are duplicates of other element names, both of which can cause odd problems.

Nothing used in the rules has an alias attached to it. Some of the other dims do but they are shared with many other cubes and i'm confident of no overlap or duplicates.

Are you editing the rules via the Rule Editor or Rule Worksheet?
If it's the Rule Worksheet try using the rule editor. Just add a space to make it save.
If using the Rule Editor, try a select all then cut and paste in to Notepad then select all and paste back again. That can help to get rid of any lurking control characters.

rule editor, have never used rule worksheets because i don't trust Excel Did also try repasting but made no difference.

Try deleting the rules, and re-entering using the rules wizard - that has been known to cure odd bugs in the past.

Didn't get that far but will bare it in mind in the future

[paulsimon]

Steve

Is there any reason why

Code: Select all

SKIPCHECK;

['attrition percentage'] = ['Attrition Count'] \ DB('Attrition','Total Actuals',!Organisation Breakdown,!Resource Breakdown,!Nomination,!Site,!Resource Source,!MonthYear,[b]'Employee Count'[/b]);

FEEDERS;

['Attrition Count'] => ['attrition percentage'];

Couldn't just be as follows?

Code: Select all

SKIPCHECK;

['attrition percentage'] = ['Attrition Count'] \ ['Total Actuals','Employee Count'] ;

FEEDERS;

['Attrition Count'] => ['attrition percentage'];

Presumably you only have Attrition Count against the three types of Attrition?

Regards

Paul Simon

[Steve Vincent]

probably not! it's just a habit of mine, if i ever need to grab data from a different part of the cube i've always used the DB functionality. Guess its just the way i think about what its doing - i find it easier to read the rule that way but either should suffice