Phillip Alday

I think it's time to emphasize that the formula syntax is in a very real sense **not** Julia code: it's a DSL and that's why it's wrapped in a macro...

> To be clear the problem is not that it is not like normal Julia code. > The problem is that * is the single most confusion choice of character...

@nalimilan the current doctests failures are unrelated to this PR, but do you have any objections to me updating them? For now, I would update update the expected output, but...

> Would you feel like adding tests for all other model families? I'm always uncomfortable when what we test varies from one family to the next. Of course -- I've...

The Windows failures look unrelated and just due to some connection flakiness. @randy3k could you restart them?

I have a few ideas about how to nudge the optimizer here, but probably won't have the time to test them / write them up for a few days. As...

@liamlundy The method error you ran into is the result of a bugfix in Julia's LinearAlgebra standard library that led to a return type changing in Julia 1.6 compared to...

I want to reiterate and expand upon a few points made by @dmbates. 1. Your weights span 7 orders of magnitude, which makes the problem numerically extremely difficult and susceptible...

@yuexiao-marketingattribution I'm a little bit surprised by the performance you've found -- 11-30 minutes is an order of magnitude slower than what I found on any my personal devices. Can...

@kliegl I don't know if MixedModels 1.x will run on Julia 1.6+. There was a change in the LinearAlgebra stdlib in Julia 1.6 that changed the type of a particular...