TM1 Forum

Hello,

My simple feeder is not working, and I don't have any idea why.

Source cube: [Atsargos]

['VISO papildymas'] = n:45;
FEEDERS;
['VISO papildymas'] =>
DB('Pirkimai', !2_DK, !2_Tarnybos_Verslai, !3_Pirkimo būdai, !4 Laikotarpiai, !5 Versijos, 'Planas sub 1');

Target cube: [Pirkimai]

SKIPCHECK;
['Planas sub 1'] = N:
DB('Atsargos', !2_DK, !2_Tarnybos_Verslai, !5 Versijos, !3_Pirkimo būdai, !4 Laikotarpiai, 'VISO papildymas');


In leaf cells I get values '45', but there is no consolidation, and when I comment out SKIPCHECK I can see consolidated values. What could be wrong with my feeder? Dimensions are mapped correctly.
-------------------------------------------------------------
Also (just out of curiosity, not related to my problem), if I write rule SKIPCHECK; ['Planas sub 1'] = N: 1; what intercube feeder should be written to turn on consolidation for this element?

For 2 different cubes to feed the contents for each of the dimensions need to be the same (Even if the names are different.) Are you sure that the content of 3_Pirkimo būdai and 5 Versijos is in line?

A feeder must have real data entry somewhere! You cant feed from 'VISO papildymas' because it is derived from a rule that is NOT FED.

(You can feed from a rule calculated value to another rule calculated value and so on but you have to start somewhere with entered "physical" data.)

Jim,

These are dimensions of my cubes:
Pirkimai: 2_DK, 2_Tarnybos_Verslai, 3_Pirkimo būdai, 4 Laikotarpiai, 5 Versijos, 5 Prognozė
Atsargos: 2_DK, 2_Tarnybos_Verslai, 5 Versijos, 3_Pirkimo būdai, 4 Laikotarpiai, Atsargos

I think DB() are mapped correctly.

lotsaram,
I had no idea that feeders need real data to function. Thank you for this information.

kudzis wrote:I had no idea that feeders need real data to function. Thank you for this information.

Then what exactly was it that you thought feeders did?

lotsaram wrote:

kudzis wrote:I had no idea that feeders need real data to function. Thank you for this information.

Then what exactly was it that you thought feeders did?

I thought feeders were working on real data and calculated values with no real data. For example, I was sure that ['A'] = N: ['B'] + 15; was very easy to feed; it turns out it's impossible.

Think ['B'] => ['A'] would work quite well

David Usherwood wrote:Think ['B'] => ['A'] would work quite well

Not if ['B'] is zero...

lotsaram wrote:A feeder must have real data entry somewhere! You cant feed from 'VISO papildymas' because it is derived from a rule that is NOT FED.

(You can feed from a rule calculated value to another rule calculated value and so on but you have to start somewhere with entered "physical" data.)

The answer is in lotsaram's post: you need to have physical data to feed from.
If your requirement really is that all (!) cells on N: level against ['VISO papildymas'] should get value 45, I would suggest to create a process that loads the 45 number in these cells. If the 45 should be in just a number of cells on N: level against ['VISO papildymas'] (which sounds more logical to me), you should think about whether there is a driver area determining what cells should get the value. Then use this driver area as your feeder.
BTW, IMO hardcoding values in rules (apart from f.i. dividing by 100 in a percentage calculation) is bad practise - I would advise to store these values in a parameter cube.

Michel

@Michel, fair point.

Michel,

static amount (45) is not in my business logic, i've used it as an example only. Basically, I'll have to create proper feeders in my source cube if I want everything working properly. Since my source cube is small and fast, I was lazy and wanted to use feeders only in target cubes.