Table 135. MLingreg (Multiple Linear Regression) Function
Field | Description |
|---|---|
Description | Returns a predicted value for a column expression based on a least squares multiple linear regression of the previous width-1 based on the sort_expression column values. When there are fewer than width-1 preceding rows, the MLinreg function computes the regression using all the preceding rows. |
Syntax: | MLingreg(value_expression, width, sort_expression_list) |
value_expression | The value expression represents a scalar numeric column expression for which a moving average is to be computed. The expression cannot contain any OLAP or aggregate functions. |
width | The width represents the number of previous rows to be used in computing the moving average. The width-1 previous rows are used to compute the linear regression and the row value itself is used to calculate the predicted value. The width value is always a positive integer constant greater than 1. The maximum width is 4096. |
sort_expression_list | The sort_expression_list is a column reference used to sort the values and to define the dependent variable for calculating the linear regression. The sort_expression_list is an expression with optional sort direction specification. The default sort direction is ascending (ASC). Only one sort_expression is allowed with this function. |