In a Spotfire linear regression model, where are the coefficients for each parameter displayed?

Product: Spotfire Analyst
Versions:All

Summary:
In a Spotfire linear regression model, the coefficients for the explanatory variables (parameters) are shown in the "Estimate" column of the "Table of Coefficients" display.

Details:
In a Spotfire linear regression model built by selecting "Tools > Regression Model... > Linear Regression" from the installed Spotfire desktop client's main menu, where are the coefficients for each parameter displayed?

Resolution:
The coefficients for the explanatory variables (parameters) are shown in the "Estimate" column of the "Table of Coefficients" display.



This is shown in the example DXP file attached to this article: "fuelframe - with regression analysis.dxp".

It is also shown in the attached "fuelframe - with regression analysis.png" screenshot image.


The DXP file contains a linear regression model built by selecting "Tools > Regression Model... > Linear Regression" from the Spotfire Analyst desktop client's main menu, then making selections that build the following model:

Mileage ~ Weight + Disp. + Fuel + Type

In this model, the [Mileage] column is the dependent variable, and the other columns are explanatory variables (parameters).

In the "Table of Coefficients" display on its page named "MyModelA", the "Estimate" column shows the regression model's estimated values for its intercept and the slope coefficient for each of its model parameters.



The [Weight], [Disp.] and [Fuel] columns (parameters) are all numeric variables.

The "FuelFrame" data table's [Type] column is a categorical variable. That is why [Type] has a separate coefficient calculated for each of its discrete categories.


Additional Details:
```````````````````

In the background, "Tools > Regression Model... > Linear Regression" generates its result via a built-in Spotfire data function that sends the data and the model definition to a call to lm() in the Spotfire desktop client's built-in local TIBCO Enterprise Runtime for R (TERR) engine.



The example shown in the "fuelframe - with regression analysis.dxp" example DXP file uses a copy of the TERR engine's built-in 'fuel.frame' example data set.


As shown in the following TERR Console session, the values in the example DXP file's "Estimate" column are the same as the coefficients calculated by lm() in TERR:

=====
TIBCO Software Inc. Confidential Information
Copyright (C) 2011-2020 TIBCO Software Inc. ALL RIGHTS RESERVED
TIBCO Enterprise Runtime for R version 6.0.0 for Microsoft Windows 64-bit

Type 'help()' for help.
Type 'q()' to quit.
>
>
>
> ffA <- Sdatasets::fuel.frame
>
>
> ffA.lm <- lm(Mileage ~ Weight + Disp. + Fuel + Type, data = ffA)
>
>
> ffA.lm

Call:
lm(formula = Mileage ~ Weight + Disp. + Fuel + Type, data = ffA)

Coefficients:

(Intercept)      Weight       Disp.        Fuel   TypeLarge  TypeMedium
. 54.200207   -0.001404    0.015372   -6.781595    0.692237    0.558911

. TypeSmall  TypeSporty     TypeVan
.  0.673369    0.046662    3.101273
>
>
=====