ASCIIOUTPUT-TI process problem

[dilip]

I have used ASCIIOUTPUT('E:\ABC.csv',v1,v2,v3,v4,numbertostring(v5)); I am using cubeview as datasource. v5 is my numeric value coming from cube.But Iam not able to see file named ABC.csv in my E:\drive nor any data inside.Rule is written in that cube which Iam using as source as cubeview. Ihave access to my E: drive. and this Ascii code I have tried written in DATA as well as Metadata part.I want all the records should be displayed in my csv.

Please suggest.

[Wim Gielis]

Does Turbo Integrator enter the data source (view)?
That is, maybe you try to export cells calculated by a rule, but you must know that by default,
TM1 skips those cells. You must indicate in the Prolog that you don't want to skip those cells.

Code: Select all

ViewExtractSkipRuleValuesSet(CUBENAME,VIEWNAME,0);

[dilip]

I have not craeted a view through TI process instead I have taken it from the cube view itself. only in metadata I have written the whole code.Even I ahve commented all the rule driven cells then also I am not seeing the the elements and values in my csv file even iam not seeing the file name exist by that name i.e ABC.csv.

please suggest some other methodif you know I want all the records in my csv file.

[declanr]

I have not craeted a view through TI process instead I have taken it from the cube view itself. only in metadata I have written the whole code.Even I ahve commented all the rule driven cells then also I am not seeing the the elements and values in my csv file even iam not seeing the file name exist by that name i.e ABC.csv.

please suggest some other methodif you know I want all the records in my csv file.

I am a bit lost as to what you are trying to say here but just to point out another possibility, is the E drive actually on your server or is it your local E drive?

[dilip]

its in my e: drive local

[declanr]

its in my e: drive local

TI Processes are running on the TM1 server.

[dilip]

NOw AsCIIOutput('E:\ABC.csv' ,'HHHHHHHHH');.. I have tried with hardcoding the value and seeing whether it is coming or not but Only when Iam writing in the prologg part the hardcoded value is coming when Iam writing in Metadata and data part That file ABC.csv is not generating in that E:\drive path.

[Steve Rowe]

If your TI is running without error then the view you are using as a source is not generating any records so the metadata tab is not running at all.
To create a view by hand right click on the cube concern and create an Export as Text view, save the view but don't export. (Suggest you read the help on using this dialogue box).
Then point your TI at this view.
You will then get an error I expect as the TM1 server cannot see your E drive, the path you specify needs to make sense from the servers view point.

[qml]

when Iam writing in Metadata and data part That file ABC.csv is not generating in that E:\drive path.

That strongly suggests that the cube view you are using as the data source is empty and so your Metadata and Data code is not being executed at all. Please make sure that you define your data source properly. Think about what sorts of values you are expecting. Are they consolidated values or leaf-level values or both? Do you want to include empty/zero values? Do you want to include rule-derived values? Think about these questions first and make sure you build your source view so that it actually returns rows.

Some useful functions to look at when defining your view:
ViewExtractSkipZeroesSet()
ViewExtractSkipCalcsSet()
ViewExtractSkipRuleValuesSet()

PS: Steve beat me to it, but if have to lose to someone, there is no shame in being beaten by someone like him.

As a final note - when putting together a question, please make sure you give as much clearly described detail as possible. People will be much more keen to help you when they understand the problem and don't have to go deciphering, guessing and asking more questions. A well formulated problem is halfway solved.

[Wim Gielis]

Some useful functions to look at when defining your view:
ViewExtractSkipZeroesSet()
ViewExtractSkipCalcsSet()
ViewExtractSkipRuleValuesSet()

PS: Steve beat me to it, but if have to lose to someone, there is no shame in being beaten by someone like him.

I guess my suggestion above wasn't clear enough to the OP Hopefully it is now that you provided more content.

[dilip]

.Thank You very much.YES i GOT IT.... fINALLY i HAVE CREATED A VIEW BY USING TI process created a subsets and assign the subsets in a view ..........now it is working fine.

But only problem is that so many records are there due to which it is unable to open the excel and text file. not all the records are captured in my text file and excel file....

Can you please suggest.

[Wim Gielis]

.Thank You very much.YES i GOT IT.... fINALLY i HAVE CREATED A VIEW BY USING TI process created a subsets and assign the subsets in a view ..........now it is working fine.

But only problem is that so many records are there due to which it is unable to open the excel and text file. not all the records are captured in my text file and excel file....

Do you skip zeros and consolidations in the exported view? Your rules are correct?
Do you want to filter in the cube and as such, limit the output?
When exporting the data out of a cube, in most cases n-level cells suffice. Hence, create an MDX statement to gather in the subset only n-level cells.

Opening a file in Notepad is also an option (or is that what you meant above ?)

[dilip]

Thank you very much.

[qml]

Hence, create an MDX statement to gather in the subset only n-level cells.

That is probably massive overkill for the rather simple task being so laboriously solved in this thread. I would say ViewExtractSkipCalcsSet is a better way to achieve the same result for at least three reasons:
a) simplicity
b) performance
c) I suspect the OP has little to no idea what you mean by the above.

[Wim Gielis]

Hello Kamil

About your 3 points above:

c) Most probably true
a) Certainly true
b) Do you have evidence that this is faster? I do not know whether it makes any difference.
What I tested in the past (very simply, I admit) was that MDX-driven subsets in a view are not slower than static subsets (loops to fill subsets with level-n)

Wim

[qml]

b) Do you have evidence that this is faster? I do not know whether it makes any difference.
What I tested in the past (very simply, I admit) was that MDX-driven subsets in a view are not slower than static subsets (loops to fill subsets with level-n)

Evidence? Only if as evidence you accept anecdotal evidence. Which no, you shouldn't.

Depending on the version of TM1, Dynamic Subsets can (but don't have to) have two main problems: constant recalculation with every element retrieval and creating locks on dimensions. Both of these may lead to the mythical 'performance issues'.

Many developers avoid MDX subsets like the plague, which, I must say, I personally disagree with. I think they have their uses, even in data sources. What I tend to do though is implicitly convert them to static subsets before use by employing the trick of adding and deleting a dummy element. That way I still get the flexibility and ease of use, and I get rid of any potential issues - which, admittedly, I have not tested with every version of the software and in every circumstance, so maybe I am being too careful purely based on some bad experiences from years ago and hearsay.

[dilip]

Thank you Wim and Steve..........