unfold

unfold copied to clipboard

Since the colorbar is scaled after the largest predictor in the (non-expanded) DM, often most of the columns just show white (i.e. values nearer to zero which is white). Maybe...

behinger

find away to compare models efficiently

7

Modelcomparisons are often necessary. How to do them in a computational efficient way with least amount of assumptions. Proposals: - R^2 - Square Error - likelihood ratios - some kind...

behinger

There should be simple functions that do the following: y = Xb + e \hat(y) = Xb 1.) Recover modeled data \hat(y) 2.) Recover residuals e 3.) Recover modeled data...

behinger

Bit of ramble & whiteboard ====== Interactions are inherently difficult to understand. We currently only have th option to add the intercept to all effects. This helps with 1xN designs....

behinger