The matrix export defaults are set on the Export Defaults tab.
Table variables may be coded in the export data file as numbers or as the label text, with or without quotes. The code equivalents will appear in the export dictionary file. Choose the coding method that will be the most compatible with your target software package. The Numeric Representation option is disabled for Case Listing matrices.
|
Numeric Representation |
The groupings of each variable will be recoded to numbers, starting with zero, in the order in which the groupings are listed in the variable's definition. For example, if the variable were "Sex" with the values in the order "Male and female”, "Male”, and "Female”, then the recoding would be: 0 = "Male and female", 1 = "Male”, and 2 = "Female”. If the order of the variable groupings in the database dictionary were changed to "Male", "Female”, and "Male and female”, then the recoding would be: 0 = "Male”, 1 = "Female”, and 2 = "Male and female”. |
|
Labels Enclosed in Quotes |
The labels of the variable groupings and values will be the same as the text shown on your matrix, with double quotes surrounding them. For a Case Listing matrix, Unformatted labels will not appear in quotes in the data file. |
|
Labels without Quotes |
The labels of the variable groupings and values will be the same as the text shown on your matrix, without double quotes surrounding them, unless you have selected the Enclose Fields Containing Delimiter in Quotes option. |
Line breaks in the data file are significant; depending on the type of matrix, they may indicate the end of a row of data, or they may separate different sets of statistics. DOS/Windows and UNIX use different character sequences to represent a line break. Choose the option that matches the platform on which you will be using your exported data file.
|
DOS/Windows (CR/LF) |
DOS/Windows expects both a carriage return (CR) and a line feed (LF). |
|
UNIX (LF) |
UNIX expects only a line feed (LF). |
When a statistic cannot be calculated in an analysis, the cell of the matrix is marked with a footnote. No number appears in the cell. Choose a missing character to represent the empty cell.
|
Space |
A space is placed wherever the value of a missing cell would appear. If you have chosen a space as the field delimiter, you may not want to choose a space as the missing character. |
|
Period |
A period is placed wherever the value of a missing cell would appear. You should consider if your software package will handle a period as a numeric missing rather than as text. |
Export data files are in a delimited format; that is, each data item (field) is separated by a certain character. Select the character to be used as the field delimiter from this list. Tab, space, comma, and semi-colon are allowed. Variables that are output as labels will likely include spaces and commas, and may include semicolons. Therefore, when choosing the character to be used as the field delimiter, follow these guidelines:
Tab is the recommended delimiter, since tab characters never appear within a field. Most commercial software will accept tab-delimited data. However, if your software does not support tab-delimited data, then the semi-colon is recommended, since it is less likely to appear in a data field than a space or comma.
If space, comma, or semi-colon is your field delimiter and you choose to output the variables as labels, you should either select Labels Enclosed in Quotes or check the Enclose Fields Containing Delimiter in Quotes option. Either of these will ensure that a data field that has a space, comma, or semi-colon in its text will be enclosed in quotes, and therefore will not be mistaken as multiple fields.
If you select comma as the field delimiter, then it is recommended that you do not use comma as the missing character.
If you select comma as the field delimiter, then it is recommended that you Remove All Thousands Separators from the numeric fields. This will ensure that a field containing the value 1,500 is stored as 1500, so that it is not mistaken as two fields with values equal to 1 and 500.
This option will place double quotes around any field value which contains the field delimiter character. For example, if you selected the comma for your field delimiter, the label All Sites will not appear with quotes in your data file, but "Uterus, NOS" will because it contains a comma.
This option removes all commas acting as thousands separators from the calculated statistics. You may wish to select this option, even if you are using a field delimiter other than the comma. The absence of the thousands separators may be easier for your software package to handle.
When checked, this option removes all footnote characters from the data file. Footnote characters can indicate such things as a positive significance test or the reason for an empty cell. When deciding whether to select this option, you should consider if the footnote information will be helpful in your analysis of the data file. This option is disabled when working with a Case Listing matrix.
This option lists the variable names before the data in the export dictionary file. The variable names are always in quotes. This can be useful when you import/copy data into another software package that uses field names.
When exporting a Frequency Matrix or a Rate Matrix, this option can be used to output the data using the same columns that appear in the matrix. Normally all the table variables appear first and the statistics are output in columns.
Users have the option of creating SAS code to read exported matrix data. The input file specified in the resulting SAS code can be provided as a relative path, or as a fully specified path.