Parallel Regression
See Also
This menu can be used to run a large number of regression models in parallel for a set of units which have multiple measurements/y-variates on each unit. The data can either be in stacked format, where all the y-variates are stacked into one, indexed by unit and y-variate, or else a set of variates for each unit which are pointed to by a pointer. The results on each y-variate are saved, indexed by the y-variate.

The analysis for the menu is performed using the MAREGRESSION procedure which performs a regression to analyse the y-values for each measurement.

The results can be saved into structures by specifying the identifier names using the store dialog which can be opened by clicking on the Store button.

Available Data

This lists data structures appropriate for the edit box which currently has focus. You can double-click a name to enter it in the edit box.

Data Format

The data can be supplied in either of the following formats:
The spreadsheet stack and unstack menus can be used to reorganise the data between these two formats.

Y-Variate

A variate containing the stacked y-variates to be analysed, or a pointer to a set of variates containing the y-variate values for each unit. The variates must all contain the data in the same order.

Units Factor

The factor that identifies the units. If the data are in the pointer format, then this should have just one entry per unit. If the data are in a single variate (stacked) format, then this factor indexes the units in the y-variate.

Y-Variates Factor

The factor that identifies the individual y-variates. If the data are in pointer format, then this should contain the information for a single unit, and all the y-variates for each unit are expected in this common order. If the data are in a single variate (stacked) format, then this factor indexes the individual y-variates in the stacked y-variate.

Unit Order Validation

A factor the same length as the factors within the regression model which indexes the units. This must have the same levels/labels as the units factor, and is used to verify that the model terms are in the same order as the data specified in the y-variate. Supplying this factor is optional, but highly recommended to validate whether the data and treatments match as expected. If the labels of the slides and check factor match, but are in a different order, the model terms will be sorted into the correct order with a warning.

Regression Model

A formula specifying the combinations of factors and variates describes the regression model for the series of slides. For a simple linear regression this will just be the name of the variate that specifies the level of independent variable used on each slide. See the page on model formulae for more details on how to specify regression models.

Weights

A variate specifying the weights for each of the units in the regression. If field is left blank then equal weights of 1 will be used.

Offset

Variate holding values to be used as an offset on the linear predictor scale. An offset is used to take account of a fixed contribution to the linear effects for each unit. If no offset is required then this field should be left blank.

Operators

This provides a quick way of entering operators in the regression model formula. Double-click on the required symbol to copy it to the current input field. You can also type in operators directly. See model formula for a description of each operator.

Action Buttons

RunRun the analysis.
CancelClose the menu without further changes.
OptionsOpens a dialog where additional options and settings can be specified for the analysis.
DefaultsSet the menu settings back to the default settings. Clicking the right mouse on this button produces a pop-up menu where you can choose to set the menu using the currently stored defaults or the GenStat default settings.
StoreOpens a dialog to specify names of structures to store the results from the analysis. The names to save the structures should be supplied before running the analysis.

See Also