[DDI-SRG] Questions on inline ncube example

[Joachim Wackerow]

Wendy,

Thank you. I have still open issues with that, see below.

Regarding examples, see the URL's at the end of the email. I'm thinking 
of this kind of tables for an exchange of structured ncubes between 
applications.

Attribute is just a string, so everything can stay there. This is no 
machine-actionable description for the fact, if the table is weighted or 
not, and no reference to a weight variable can be made (the inclusion of 
a weight variable can make sense in a structured ncube/table). Regarding 
the type of the measure, see also number 7 below.

For using nested coding schemes a virtual code for the margin sums would 
be necessary, which doesn't exist. Is this really a good solution? An 
application can not detect that this virtual code is an margin sum. My 
approach with the two-dimensional tables (n-1) makes the assumption that 
the three-dimensional table and the two-dimensional tables have a 
relationship, which is again for an application perhaps not easy to detect.

Achim

Wendy Thomas wrote:
> 
> send examples of the table contents you want to describe. On first shot 
> my responses would be:
> 
> no
> yes (using the attribute)
> yes (nested coding schemes
> 
> On Wed, 28 Nov 2007, Joachim Wackerow wrote:
> 
>> Some additional questions on this:
>>
>> How is it possible to describe a sparse matrix in an inline ncube
>> without including all the empty cells?
>>
>> Can a weighted table described as such? Or did I miss something?
>>
>> Re 5: The margin sums can be described for a three-dimensional table by
>> three additional two-dimensional ncubes. Is there an easier way?
>>
>> Achim
>>
>> Joachim Wackerow wrote:
>>> Wendy,
>>>
>>> I'm looking into how to use ncubes for the exchange of tables (structure
>>> approach). I made a inline ncube example for a three-dimensional table.
>>> Please look over it if everything is correct. I'm not sure with this.
>>>
>>> Making this example, some questions raised:
>>>
>>> 1. 'NCubeInstanceReference' should reference the NCube description in
>>> the logical product. But a element 'NCubeInstance' exists in the ncube
>>> module. This is confusing to me. Assuming the documentation is correct,
>>> then the name should not express that it is a reference to a
>>> NCubeInstance but to a Ncube like 'NCubeReference'.
>>>
>>> 2. Repeating the container 'MeasureValue' for each cell, also the
>>> 'MeasureReference' is repeated in this case of a regular
>>> three-dimensional table. Should there exist a 'DefaultMeasureReference'?
>>>
>>> 3. Ordering the MeasureValue's I assumed a hierarchy in the dimensions
>>> expressed by the rank, the dimension with the first rank changing the
>>> fastest, the dimension with the second rank changing the second fastest
>>> (the sequence of the vlues corresponds to the cells in the display table
>>> from left to right in one row, then the next row, then the next
>>> sub-table). Is this correct? I didn't find anything related in the
>>> documentation.
>>>
>>> 4. Does 'GrossRecordStructure' make sense with a inline ncube (with
>>> 'LogicalRecordReference' and 'PhysicalRecordSegment')? It is required 
>>> here.
>>>
>>> 5. How would you describe the margin sums? It can make sense for large
>>> tables.
>>>
>>> 6. I would like to describe the same table with percentages. I assume
>>> that the 'RecordLayout' must be repeated with the percentage values and
>>> a reference to a measure variable for percentages in logicalproduct.
>>>
>>> 7. The type of the measure is described by the label of the measure
>>> variable in logicalproduct or by the related 'ConceptReference'. I think
>>> it would make sense to describe it by a controlled vocabulary like for
>>> category statistics.
>>>
>>> 8. The concept of 'NCube/Attribute' is not clear to me. Can you please
>>> describe it.
>>>
>>> The display tables are available at:
>>> values
>>> http://exanda.zuma-mannheim.de/Study_Wohlfahrtssurvey/Independent_SCHULAB/Dependent_ALTER1/Group_GESCHL/Weight_Weight/OutputType_Value/FileType_HTML/Trivariate.html 
>>>
>>>
>>> row percentages
>>> http://exanda.zuma-mannheim.de/Study_Wohlfahrtssurvey/Independent_SCHULAB/Dependent_ALTER1/Group_GESCHL/Weight_Weight/OutputType_RowPercent/FileType_HTML/Trivariate.html 
>>>
>>>
>>>
>>> Achim
>>>
>>>
>>> ------------------------------------------------------------------------
>>>
>>> _______________________________________________
>>> DDI-SRG mailing list
>>> DDI-SRG at icpsr.umich.edu
>>> http://www.icpsr.umich.edu/mailman/listinfo/ddi-srg
>>
>> _______________________________________________
>> DDI-SRG mailing list
>> DDI-SRG at icpsr.umich.edu
>> http://www.icpsr.umich.edu/mailman/listinfo/ddi-srg
>>
> 
> Wendy L. Thomas                          Phone: +1 612.624.4389
> Data Access Core Director         Fax:   +1 612.626.8375
> Minnesota Population Center              Email: wlt at pop.umn.edu
> University of Minnesota
> 50 Willey Hall
> 225 19th Avenue South
> Minneapolis, MN 55455