Closed leespen1 closed 2 years ago
Thanks for fixing it. This routine has not been used recently, but could be a good entry point for your higher order scheme.
No problem! I agree, for my current goals it's nice to have a function which only evolves the state vector.
Get Outlook for Androidhttps://aka.ms/ghei36
From: Anders Petersson @.> Sent: Wednesday, July 6, 2022 7:25:45 PM To: LLNL/Juqbox.jl @.> Cc: Lee, Spencer @.>; Author @.> Subject: Re: [LLNL/Juqbox.jl] Fixed missing reference to linear_solver (PR #26)
Thanks for fixing it. This routine has not been used recently, but could be a good entry point for your higher order scheme.
— Reply to this email directly, view it on GitHubhttps://urldefense.com/v3/__https://github.com/LLNL/Juqbox.jl/pull/26*issuecomment-1176852081__;Iw!!HXCxUKc!2I-yKa2OxXzaOHCFJTRQR1KGFmMcjcixP9NCFJ4ZsFcdJJCzUeRQLgMiNDQxzxnwXmCt1aqwusBw6gwUNIfTrMkQ$, or unsubscribehttps://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AM6ZUBDAFYQOJOJYEIDDLNLVSYIXTANCNFSM523KC4DQ__;!!HXCxUKc!2I-yKa2OxXzaOHCFJTRQR1KGFmMcjcixP9NCFJ4ZsFcdJJCzUeRQLgMiNDQxzxnwXmCt1aqwusBw6gwUNGUAgT_o$. You are receiving this because you authored the thread.Message ID: @.***>
In file
src/evalobjgrad.jl
, functioneval_forward
, line 1982 made reference to missing objectlinear_solver
, which is not created anywhere in the function (and is not a global variable), causing an error at runtime. The similar functiontraceobjgrad
contains the linelinear_solver = params.linear_solver
, so I copied that line intoeval_forward
(in a similar position).After making the change, I no longer get a runtime error when running
eval_forward
.