Add factorial smc

[SamDuffield]

Adds SMC factorializer code.

Adaptive resampling doesn't play nice with factorial so we assume resampling is always applied (i.e. no weights carried through). This could probably be documented better.

[SamDuffield]

Ah actually cuthbert SMC applies resampling at the start of the filter_combine. This might be problematic, will think about it more

[AdrienCorenflos]

I'm not willing to change that one lol. Monte Carlo outputs should come weighted rather than resampled to reduce variance.

…

On Thu, 21 May 2026, 17:21 SamDuffield, ***@***.***> wrote: *SamDuffield* left a comment (state-space-models/cuthbert#240) <#240 (comment)> Ah actually cuthbert SMC applies resampling at the start of the filter_combine. This might be problematic, will think about it more — Reply to this email directly, view it on GitHub <#240 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AEYGFZ5NGLIGOBAOBAE4TST434UI3AVCNFSM6AAAAACZH5WSRGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DKMJQGI3DSNJWGU> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[SamDuffield]

I'm not willing to change that one lol. Monte Carlo outputs should come weighted rather than resampled to reduce variance.

agreed! Will try find a solution at the factorial level