Fix FluidTransient<Slice> allocations #290

weefuzzy · 2024-12-18T18:09:12Z

Fixes #270

Ensure that clients instantiate the transient algorithms with actually sensible max values for order, blocksize, padding
Add some extra asserts to make future us have an easier time tracing this stuff
Properly finish the job for ARModel and ensure that the model order no longer assumed to be the size of the parameters vector 😬

– decouple current order from mParameters.size() – add some asserts – delete old code

tremblap

I've tested in SC with real time changes of all parameters, and UTs, and it is stable now, so I reckon we should merge before 1.0.8 - I am testing now in Linux to see how it behaves.

weefuzzy added 4 commits December 18, 2024 17:55

TransientClient: Proper limits for algorithm instantiation

64fa0b7

TransientExtraction: enforce invariants in constructor

378b3c1

ARModel: Fix and tidy

6f16021

– decouple current order from mParameters.size() – add some asserts – delete old code

ARModel: Formatting

aa01f2c

weefuzzy added the bug Something isn't working label Dec 18, 2024

weefuzzy requested review from AlexHarker and tremblap December 18, 2024 18:09

weefuzzy self-assigned this Dec 18, 2024

weefuzzy mentioned this pull request Dec 18, 2024

fix #270 - memory allocation for the autocorrelation in transient* #288

Closed

proposed change to max pad size

b3cf21f

tremblap approved these changes Dec 19, 2024

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Fix FluidTransient<Slice> allocations #290

Fix FluidTransient<Slice> allocations #290

weefuzzy commented Dec 18, 2024

tremblap left a comment

Fix FluidTransient<Slice> allocations #290

Are you sure you want to change the base?

Fix FluidTransient<Slice> allocations #290

Conversation

weefuzzy commented Dec 18, 2024

tremblap left a comment

Choose a reason for hiding this comment