Seth Axen

> I think the point of `test_rrule()` is to make all possible checks unless it's instructed otherwise. This is really useful when you develop a new `rrule`, but causes unnecessary...

I'm only seeing Cholesky-related errors. I tested locally that these are resolved by https://github.com/JuliaDiff/ChainRules.jl/pull/633.

Julia's `evalpoly` implementation is equivalent to one that uses `foldr`, but that actually has worse performance for some reason. ```julia julia> _evalpoly(x, p) = foldr((pi, y) -> muladd(x, y, pi),...

> Yes, I hadn't tried but individual sparse vectors still seem like a lot of overhead, especially if (as is common) you are covering the whole array. Did you literally...

> Compared to about 100ms for the `sparse` solution, on my computer, and 118.672 μs for the `foldl` one! I'm not precisely sure when this gets accumulated, and whether this...

> I'm curious why/if `Buffer` doesn't do the job here. It seems like it should solve the main problem – huge number of array allocations – quite neatly, even if...

> I'm curious why/if `Buffer` doesn't do the job here. It seems like it should solve the main problem – huge number of array allocations – quite neatly, even if...

Ah okay, so this has the same solution as https://github.com/JuliaPy/PyCall.jl/issues/722#issuecomment-575982875, namely ```bash $ LD_PRELOAD=$HOME/.julia/conda/3/lib/libstdc++.so julia _ _ _ _(_)_ | Documentation: https://docs.julialang.org (_) | (_) (_) | _ _ _|...

Could you provide a minimal example of a Python class for which `dir` fails but `getmember` works? See for example the [test class](https://github.com/JuliaPy/PyCall.jl/blob/8a98fb45ef39d09e11c1270e9120a32df2578d50/test/runtests.jl#L298-L300) added in #879 for which `dir` works...

Yes, for Pathfinder we need the inverse Hessian approximations for every step in the trajectory when using L-BFGS, and we reconstruct this from the trajectory and gradients (see https://github.com/sethaxen/Pathfinder.jl/blob/8c2d00e2ea3612db814390fbddcb2f915d360156/src/inverse_hessian.jl). If...