dallan-keylogic

@adowling2 From the solver trace @shermanjasonaf posted, it looks like all constraints are solved to within `1e-6`, so I don't think primal infeasibility is the driver here. An added complication...

Alright, since it looks like the consensus choice is to discretize domains early, that has a couple of implications. 1. Fixing variables must always be viewed as a fragile operation....

I've pretty much decided that we need to discretize the time domain before anyone starts fixing variables. ~~This bears onto the current heat exchanger draft PR~~ (Edit: Not actually true),...

Windows 11, Microsoft Edge Version 103.0.1264.49 (Official build) (64-bit). Google Chrome Version 103.0.5060.114 (Official Build) (64-bit)

What is the status of these improvements, @adowling2 ? In particular, does the current version of Degeneracy Hunter support fixed variables? I'm finding that a certain IDAES model has a...

> Check that flowsheet.report() works with distillation columns - there appears to be a bug somewhere. Found the cause of this while investigating #480. The `UnitModelBlock.add_inlet_port()` method currently adds an...

@adowling2 It's actually only around an 800x800 square problem. Using a normal SVD works fine. Using `solver="lobpcg"` didn't help in my case. We need to form the matrix X'X then...

I spent a morning trying a few things without success, then just resigned myself to using the dense SVD. However, I finally have some time to work on this again,...

@adowling2 @shermanjasonaf I pushed my first attempt at a homebrew version of `svds` to the `homebrew_svds` branch of [my IDAES fork.](https://github.com/dallan-keylogic/idaes-pse/tree/homebrew_svds) Like `svds`, we use `scipy.sparse.linalg.eigsh` on the normal matrix...

@eslickj That modification did not work.