TM1 Forum

Hi,

I have an existing TI process which doesn't return the correct value of the cube I'm referencing to. The current process doubles the expected value. ViewZeroOut function is running before the process so I'm kinda stuck why this is happening. I can't seem to find anything wrong in the process, the codes are just generated and I didn't add any since the process will just pull the data from the other cube. Can anyone please enlighten me with this.

Thank you.

Sounds like a lot is happening and without seeing your code it will be very hard to give a helpful response.

I would comment out all aspects of the process except for the ViewZeroOut to ensure that is working as expected. Then include an asciioutput to ensure the remainder of your code is aggregating/ exporting/ calculating values as expected and hopefully identify where duplications may or may not be appearing

Edward Stuart wrote: ↑Fri Oct 30, 2020 11:08 am Sounds like a lot is happening and without seeing your code it will be very hard to give a helpful response.

I would comment out all aspects of the process except for the ViewZeroOut to ensure that is working as expected. Then include an asciioutput to ensure the remainder of your code is aggregating/ exporting/ calculating values as expected and hopefully identify where duplications may or may not be appearing

Hi Edward,

I appreciate your comment. I've tried including the asciioutput and it seems that the code is aggregating/exporting correctly. But the file doesn't include the value since it is not an element.

Here is the existing code on the process:
#****Begin: Generated Statements***
vMeasure='Total_Cost';
vVersion=pVersion;
if (VALUE_IS_STRING=1, CellPutS(SVALUE,'Test_Cube',vYear,vPeriod,vVersion,vCurrency,vPlant,vComponentSKU,vMeasure), CellPutN(NVALUE, 'Test_Cube', vYear,vPeriod,vVersion,vCurrency,vPlant,vComponentSKU,vMeasure));
#****End: Generated Statements****

It is just system generated since I will only pull the data from my Source Cube.

Many Thanks,
Jaine

Add an ASCIIOUTPUT after the generated statements:
This will tell you what you are reading. Based on what you've said, I'm thinking your datasource is not what you think it is.

PavoGa wrote: ↑Tue Nov 03, 2020 5:49 pm Add an ASCIIOUTPUT after the generated statements:

Code: Select all

# to use this you'll have to define sDataLogfile with filename and path for the output.
ASCIIOutput(sDataLogfile, EXPAND('%NValue%, %vYear%, %vPeriod%, %vCurrency%, %vPlant%, %vComponentSKU%, %vMeasure%'));

This will tell you what you are reading. Based on what you've said, I'm thinking your datasource is not what you think it is.

I already tried adding this in the process and it seems that it is aggregating as expected.

Where is the ViewZeroOut? In the prolog or in a subprocess?

j.calibs wrote: ↑Wed Nov 04, 2020 12:48 am I already tried adding this in the process and it seems that it is aggregating as expected.

Focussing on the word "aggregating" here.

If your source has more than 1 record that should go to the same target cell; you need to specify that it should aggregate those values - for example swap your CellPutN for a CellIncrementN statement. Otherwise you will just end up with the last source record's value.

I have re-read the first post and realised that I somehow missed the main point that its doubling a value. This makes my last reply an unlikely cause, not impossible however if you are trying to aggregate 2 values that just so happen to include 1 being a negative 50% of the other.

Is the end result always exactly double what you want? Or if you run the process 3 times does it end up as triple? And 4 times quadruple etc?
If so, then as others have suggested your zeroout would be the issue - again you can test this by manually clearing the cube once before you run it.

Also have you drilled down to the most granular cell level to see if its doubled there? Or are you just seeing the double value at a high consolidated level? If so it is possible that n-level elements are appearing in your hierarchy twice, or have a weight of 2 where they would usually be weighted as 1.

If everything else fails, you could perhaps provide a small sample of source data and show the result it is creating? Along with providing your full code from every tab, including the zeroout statement etc.

j.calibs wrote: ↑Wed Nov 04, 2020 12:48 am

PavoGa wrote: ↑Tue Nov 03, 2020 5:49 pm Add an ASCIIOUTPUT after the generated statements:

Code: Select all

# to use this you'll have to define sDataLogfile with filename and path for the output.
ASCIIOutput(sDataLogfile, EXPAND('%NValue%, %vYear%, %vPeriod%, %vCurrency%, %vPlant%, %vComponentSKU%, %vMeasure%'));

This will tell you what you are reading. Based on what you've said, I'm thinking your datasource is not what you think it is.

I already tried adding this in the process and it seems that it is aggregating as expected.

As pointed out by Declan and what I was targeting, there is no way the process is aggregating values in a cell with a CELLPUT. Not based on the little bit of code you have shown.

j.calibs wrote: ↑Wed Nov 04, 2020 12:48 am

PavoGa wrote: ↑Tue Nov 03, 2020 5:49 pm Add an ASCIIOUTPUT after the generated statements:

Code: Select all

# to use this you'll have to define sDataLogfile with filename and path for the output.
ASCIIOutput(sDataLogfile, EXPAND('%NValue%, %vYear%, %vPeriod%, %vCurrency%, %vPlant%, %vComponentSKU%, %vMeasure%'));

This will tell you what you are reading. Based on what you've said, I'm thinking your datasource is not what you think it is.

I already tried adding this in the process and it seems that it is aggregating as expected.

As pointed out by Declan and what I was targeting, there is no way the process is aggregating values in a cell with a CELLPUT. What you are describing is not possible based on the little bit of code you have shown.

We need to see the code and some sample source data and the cube results.