DocsRegression
Analyze — Statistical Methods
Regression
Linear regression (ordinary least squares) models how a numeric target column changes with one or more predictors.
Quick Start
- 1
Select your target and predictor columns
All columns you'll assign as dependent or independent variables need to be numeric.
- 2
Choose Linear regression (OLS)
Pick a dependent variable first, then check one or more independent variables from the remaining columns.
- 3
Run and read the results
Results render inline once the job completes.
Fields
Dependent variable (Y)The target column you're trying to explain or predict. Required.Independent variables (X)One or more predictor columns. Must be numeric, and can't include the dependent variable itself.Good to know
- Picking a dependent variable first narrows the independent variable list automatically — you can't accidentally use the same column as both.
- More predictors isn't always better — check correlation between your candidate predictors first to avoid redundant variables.
- All predictors must be numeric — categorical columns need to be recoded before they can be used here.