Difference between revisions of "MdArrayToTable"
Line 5: Line 5:
  
 
Transforms a multi-dimensional array, ''A'', into a two-dimensional
 
Transforms a multi-dimensional array, ''A'', into a two-dimensional
−
array (i.e., a table) indexed by ''I'' and ''L''. The result contains one
+
array (i.e., a relational table) indexed by ''I'' and ''L''. The result contains one
 
row along ''I'' for each element of ''A''. ''L'' must contain a list of names
 
row along ''I'' for each element of ''A''. ''L'' must contain a list of names
 
of the indexes of ''A'', followed by one final element. All elements of
 
of the indexes of ''A'', followed by one final element. All elements of
 
L must be text values. The column corresponding to the final element
 
L must be text values. The column corresponding to the final element
 
of ''L'' contains the cell value. If L does not contain all the
 
of ''L'' contains the cell value. If L does not contain all the
−
indexes of ''A'', array abstraction will create a set of tables indexed
 
−
by the dimensions not listed in ''L''.
 
−
 
−
Transforms a multi-dimensional array, ''A'', into a two-dimensional
 
−
array (i.e., a table) indexed by ''I'' and ''L''. The result contains one
 
−
row along ''I'' for each element of ''A''. ''L'' must contain a list of names
 
−
of the indexes of ''A'', followed by one final element. All elements of
 
−
''L'' must be text values. The column corresponding to the final element
 
−
of ''L'' contains the cell value. If ''L'' does not contain all the
 
 
indexes of ''A'', array abstraction will create a set of tables indexed
 
indexes of ''A'', array abstraction will create a set of tables indexed
 
by the dimensions not listed in ''L''.
 
by the dimensions not listed in ''L''.
Line 26: Line 17:
 
[[MdTable]]( ) on a table that contains a row for only the nonzero
 
[[MdTable]]( ) on a table that contains a row for only the nonzero
 
elements of the array.
 
elements of the array.
 +
 +
== Creating a fact table ==
 +
 +
''New to [[What's new in Analytica 4.2?|Analytica 4.2]]''
 +
 +
::[[MdArrayToTable]](A,I,L'',valueIndex'')
 +
 +
A fact table is a generalization of a simple relational table.  In a fact table, the first ''N'' columns identify cell coordinates, and the last ''K'' columns hold measured values (often called ''measures'', ''facts'', or just ''values'').  This generalized the simple relational table where only the single final row holds a value.
 +
 +
The optional parameter, «valueColumn», specifies an index of «A» that is used for the value columns in the result.  The first [[Size]](L) - [[Size]](valueIndex) columns of «L» are then the coordinates, and the final [[Size]](valueIndex]] columns of «L» hold the values.
 +
 +
When omitting cells having no data, all values in «A» along the valueIndex must be zero or null.  If any of the values have data, the corresponding cell (i.e., row in the result) is considered to have data.
 +
 +
== Positional Coordinates ==
 +
 +
''New to [[What's new in Analytica 4.2?|Analytica 4.2]]''
 +
 +
::[[MdArrayToTable]](...,positional:true)
 +
 +
Normally, the first ''N'' columns of the result specify coordinates using index ''labels''.  When the optional parameter «positional» is specified as true, index positions are returned rather than labels.
  
 
= Library =
 
= Library =

Revision as of 23:23, 20 August 2009


MdArrayToTable(A,I,L)

Transforms a multi-dimensional array, A, into a two-dimensional array (i.e., a relational table) indexed by I and L. The result contains one row along I for each element of A. L must contain a list of names of the indexes of A, followed by one final element. All elements of L must be text values. The column corresponding to the final element of L contains the cell value. If L does not contain all the indexes of A, array abstraction will create a set of tables indexed by the dimensions not listed in L.

If the number of elements in I is equal to the number of nonzero elements of A, MDArrayToTable() acts like the inverse of MdTable( ) on a table that contains a row for only the nonzero elements of the array.

Creating a fact table

New to Analytica 4.2

MdArrayToTable(A,I,L,valueIndex)

A fact table is a generalization of a simple relational table. In a fact table, the first N columns identify cell coordinates, and the last K columns hold measured values (often called measures, facts, or just values). This generalized the simple relational table where only the single final row holds a value.

The optional parameter, «valueColumn», specifies an index of «A» that is used for the value columns in the result. The first Size(L) - Size(valueIndex) columns of «L» are then the coordinates, and the final Size(valueIndex]] columns of «L» hold the values.

When omitting cells having no data, all values in «A» along the valueIndex must be zero or null. If any of the values have data, the corresponding cell (i.e., row in the result) is considered to have data.

Positional Coordinates

New to Analytica 4.2

MdArrayToTable(...,positional:true)

Normally, the first N columns of the result specify coordinates using index labels. When the optional parameter «positional» is specified as true, index positions are returned rather than labels.

Library

Array

Example

Define these global objects (as indexes or variables): 

Rows := sequence(1,size(Cost_in_time))
Cols := [‘Mpg’,’Time’,’Car_type’,’Cost’]

MDArrayToTable(Cost_in_time,Rows,Cols) →
Rows v, Cols >>
Mpg Time Car_type Cost
1 26 0 VW 2185
2 26 0 Honda 2385
3 26 0 BMW 3185
4 26 1 VW 2294
5 26 1 Honda 2314
6 26 1 BMW 3294
7 26 2 VW 2409
...
45 35 4 BMW 5175

Note: The expression for doing the same transformation using local indexes looks like 

Index Rows := sequence(1,size(Cost_in_time))
Index Cols := [‘Mpg’,’Time’,’Car_type’,’Cost’]
MDArrayToTable(Cost_in_time,Rows,Cols)

See Also

Comments


You are not allowed to post comments.