If you want to store a subset of the units in a vector rather than restricting the original data set, you can use the **Subset** menu. This method forms a vector containing a subset of the values from another vector. You can form a subset for either a variate, factor or text. The subset is defined by a logical condition, which is specified by the matching conditions where units with the value TRUE (neither zero nor missing value) for the condition are included in the subset: others are omitted.

- From the menu select
**Data | Subset**.

The**Available data**field lists variates, factors and texts which are available to form a subset from. - In the
**Available data**field, double-click the items you want to move to the**Data to subset**field. - Type a name for your subset.
- Enter the logical condition that you want to use to subset your data.

You can combine expressions by chaining them together in the first field, or select**And**or**Or**and input another expression into the lower field.

You are not restricted to the logical operators in the**Comparisons**field, but can build your own condition using standard logical functions. - Click
**Run**to create the subset.The subset is now available for analysis and can be seen in the Data View by clicking the*Data*tab at the bottom left of the window. - To view the subset as a spreadsheet, in the Data View, click and drag your subset over any open spreadsheet then release.

This creates a new spreadsheet containing only your subset. You can now save it or run an analysis on it.

Matching condition |
Input your conditions into one or both fields. You can combine expressions by chaining them together in the first field, or select For example, the expression below, which has |

Create expression |
Click this to open the Form Expression dialog, which contains more logical operators to help build your restriction. |

Logical operators must be delimited by dots (.and. .or. .not.)

Strings must be enclosed in quotes (‘Control’, ‘Once’, ‘Twice’.)

**Example**

Match all rows where the column Drench has a value of Once or Twice, and the Lwt2 column has a value greater than 50:

Drench .in. ‘Once’ .or. Drench .in. ‘Twice’ .and. Lwt2>50

You are not restricted to the logical operators in the **Comparisons** field, but can build your own condition using standard logical functions.

**Example**

Match all rows where the difference in value between Lwt1 and Lwt2 is greater than 15:

(Lwt2-Lwt1)>15