Identifying Variables in the Data Source
After you define a data source, TurboIntegrator assigns a variable to each column in the source. You must identify these variables by type and content.
To illustrate this process, consider the following text data:
| Column 1 | Column 2 | Column 3 | Column 4 | Column 5 | Column 6 |
|---|---|---|---|---|---|
| New England | Massachusetts | Boston | Supermart | Feb | 2000000 |
| New England | Massachusetts | Springfield | Supermart | Feb | 1400000 |
| New England | Massachusetts | Worcester | Supermart | Feb | 2200000 |
| New England | Connecticut | Hartford | Supermart | Feb | 1240000 |
| New England | Connecticut | New Haven | Supermart | Feb | 2700000 |
| New England | Connecticut | Greenwich | Supermart | Feb | 1700000 |
The first 3 columns form a hierarchy for a Location dimension that you will build from the source text file:
- The New England consolidation is at the top of the hierarchy.
- The states Massachusetts and Connecticut are one level below New England.
- The third column, containing city names such as Boston and Hartford, supplies simple elements at the lowest level of the hierarchy.
- The remaining columns are not used for the creation of the Location dimension.
Here is the Variables tab on the TurboIntegrator window for this data structure:
| Variable Name | Variable Type | Sample Value |
|---|---|---|
| V1 |
String |
New England |
| V1 |
String |
New England |
| Massachusetts |
String |
Massachusetts |
| Boston |
String |
Boston |
| SuperMart |
String |
SuperMart |
| Feb |
String |
Feb |
| V6 |
Numeric |
2000000 |
TurboIntegrator assigns a variable name to each column, and assigns a variable type based on the sample value for each column.
The default variables names, such as V1 and Massachusetts, can be changed. It is good practice to give the variables a meaningful name. Having meaningful names makes the TurboIntegrator scripts easier to read and troubleshoot.
To edit a variable name, click the name in the Variable Name column and type a new name. For this exercise, the names of the first three variables were edited like this:
| Sample Value | Variable Name |
|---|---|
| New England | Region |
| Massachusetts | State |
| Boston | City |
A variable name must begin with a letter, and can contain only these characters:
| Character |
Description |
|---|---|
| Upper-case Letters |
A through Z |
| Lower-Case Letters |
a through z |
| Digits |
0 through 9 |
| Period |
. |
| Underscore |
_ |
| Dollar Sign |
$ |
The Variable Type field identifies the contents of the column. For instance, the first column of this data contains the string "New England". TurboIntegrator correctly identifies the variable type as String.
Note: The Variable Type fields are usually set accurately for ASCII data, but not for data extracted from an ODBC data source.
The Contents field can be defined with one of these settings:
| Option |
Description |
|---|---|
| Ignore |
Ignore the contents of the column when processing the data source. |
| Element |
The column contains simple elements for the dimension you want to create. |
| Consolidation |
The column contains consolidated elements for the dimension you want to create. |
| Data |
The column contains data values. For this example, you should ignore the column containing the data values. Columns containing data values are not imported when you are creating a dimension. |
| Attribute |
The column contains element attributes for the dimension you want to create. |
| Other |
The column contains data that does not fall into any of the previous four categories. Typically, this setting is used for columns containing data that will be processed through custom variables and formulas. |
The text data in this example contains elements and consolidations for a Location dimension:
- It does not contain any attributes.
- It does contain data values, but those values are irrelevant to the creation of the Location dimension, as are the elements from other dimensions.
To define the variables for the Location dimension: