# Particle Filtering with Missing Observations **URL:** \ **Category:** ML/Bayesian estimation\ **Created:** [13 June 2024 05:47 UTC](https://forum.dynare.org/t/particle-filtering-with-missing-observations/25911 "2024-06-13T05:47:43Z")\ **Posts on this page:** 10\ **Page:** 1
**Author:** ![lefarmer](https://forum.dynare.org/letter_avatar_proxy/v4/letter/l/a698b9/32.png) [@lefarmer](https://forum.dynare.org/u/lefarmer)\ **Post date:** [13 June 2024 05:47 UTC](https://forum.dynare.org/t/particle-filtering-with-missing-observations/25911/1 "2024-06-13T05:47:44Z")
I am wondering if there is any plan for supporting particle filter estimation of nonlinear models with missing observations. Is there any workaround to this right now? I currently want to perform estimation of a model at a monthly frequency but where some data are only observed at an annual or quarterly frequency. ---
**Author:** ![jpfeifer](https://forum.dynare.org/user_avatar/forum.dynare.org/jpfeifer/32/5044_2.png) [@jpfeifer](https://forum.dynare.org/u/jpfeifer)\ **Post date:** [17 June 2024 17:21 UTC](https://forum.dynare.org/t/particle-filtering-with-missing-observations/25911/2 "2024-06-17T17:21:57Z")
@stepan-a and @frederic.karame may know. But looking at the codes, it might be straightforward to implement for some filters. Which one did you have in mind? ---
**Author:** ![lefarmer](https://forum.dynare.org/letter_avatar_proxy/v4/letter/l/a698b9/32.png) [@lefarmer](https://forum.dynare.org/u/lefarmer)\ **Post date:** [17 June 2024 17:38 UTC](https://forum.dynare.org/t/particle-filtering-with-missing-observations/25911/3 "2024-06-17T17:38:22Z")
I was planning to use an auxiliary particle filter for parameter/state estimation but even a bootstrap particle filter would be fine. ---
**Author:** ![frederic.karame](https://forum.dynare.org/letter_avatar_proxy/v4/letter/f/53a042/32.png) [@frederic.karame](https://forum.dynare.org/u/frederic.karame)\ **Post date:** [18 June 2024 09:16 UTC](https://forum.dynare.org/t/particle-filtering-with-missing-observations/25911/4 "2024-06-18T09:16:29Z")
It is not difficult a priori. I will have a look Thursday. ---
**Author:** ![lefarmer](https://forum.dynare.org/letter_avatar_proxy/v4/letter/l/a698b9/32.png) [@lefarmer](https://forum.dynare.org/u/lefarmer)\ **Post date:** [18 June 2024 12:23 UTC](https://forum.dynare.org/t/particle-filtering-with-missing-observations/25911/5 "2024-06-18T12:23:23Z")
Thank you! ---
**Author:** ![frederic.karame](https://forum.dynare.org/letter_avatar_proxy/v4/letter/f/53a042/32.png) [@frederic.karame](https://forum.dynare.org/u/frederic.karame)\ **Post date:** [20 June 2024 10:09 UTC](https://forum.dynare.org/t/particle-filtering-with-missing-observations/25911/6 "2024-06-20T10:09:09Z")
Try this and let me know (put these two files in the same rep that your mod-file). [auxiliary\_particle\_filter.m](https://forum.dynare.org/uploads/short-url/ipU7RK9QpOvIAOdnaNCWKZjBPnq.m) (9.8 KB) [sequential\_importance\_particle\_filter.m](https://forum.dynare.org/uploads/short-url/wXdNnSepjsoqcMeyvMiCa3zsUGF.m) (7.9 KB) I will push them in Dynare unstable if it is OK. Best, ---
**Author:** ![lefarmer](https://forum.dynare.org/letter_avatar_proxy/v4/letter/l/a698b9/32.png) [@lefarmer](https://forum.dynare.org/u/lefarmer)\ **Post date:** [20 June 2024 13:04 UTC](https://forum.dynare.org/t/particle-filtering-with-missing-observations/25911/7 "2024-06-20T13:04:23Z")
Thank you! I’ll give this a test and let you know. ---
**Author:** ![cderesende](https://forum.dynare.org/letter_avatar_proxy/v4/letter/c/ecd19e/32.png) [@cderesende](https://forum.dynare.org/u/cderesende)\ **Post date:** [20 June 2024 19:00 UTC](https://forum.dynare.org/t/particle-filtering-with-missing-observations/25911/8 "2024-06-20T19:00:01Z")
Dear Prof. Pfeifer, I am trying for the first time to use the particle filter for the estimation of a nonlinear model with Dynare. A linear version of the mode works fine. The command I am using is: estimation(order=2, datafile=data\_Climate, number\_of\_particles=1000, mode\_compute=11, silent\_optimizer,mh\_replic=0); And the error messages I am getting are: ESTIMATION\_CHECKS: There was an error in computing the likelihood for initial parameter values. ESTIMATION\_CHECKS: If this is not a problem with the setting of options (check the error message below), ESTIMATION\_CHECKS: you should try using the calibrated version of the model as starting values. To do ESTIMATION\_CHECKS: this, add an empty estimated\_params\_init-block with use\_calibration option immediately before the estimation ESTIMATION\_CHECKS: command (and after the estimated\_params-block so that it does not get overwritten): Error using initial\_estimation\_checks Error using initial\_estimation\_checks initial\_estimation\_checks:: particle filtering requires measurement error on the observables Error in dynare\_estimation\_1 (line 165) oo\_ = initial\_estimation\_checks(objective\_function,xparam1,dataset\_,dataset\_info,M\_,estim\_params\_,options\_,bayestopt\_,bounds,oo\_); Error in dynare\_estimation (line 105) dynare\_estimation\_1(var\_list,dname); Error in Climate\_NonLin\_estim.driver (line 628) oo\_recursive\_=dynare\_estimation(var\_list\_); Error in dynare (line 310) evalin(‘base’,[fname ‘.driver’]); Not sure I understand what the problem is and how to follow the tentative instructions provided. I wonder whether anyone could offer some help. Many thanks! ---
**Author:** ![jpfeifer](https://forum.dynare.org/user_avatar/forum.dynare.org/jpfeifer/32/5044_2.png) [@jpfeifer](https://forum.dynare.org/u/jpfeifer)\ **Post date:** [21 June 2024 07:55 UTC](https://forum.dynare.org/t/particle-filtering-with-missing-observations/25911/9 "2024-06-21T07:55:02Z")
As the message says, estimation with a particle filter requires measurement error on the observables. ---
**Author:** ![cderesende](https://forum.dynare.org/letter_avatar_proxy/v4/letter/c/ecd19e/32.png) [@cderesende](https://forum.dynare.org/u/cderesende)\ **Post date:** [25 June 2024 21:25 UTC](https://forum.dynare.org/t/particle-filtering-with-missing-observations/25911/10 "2024-06-25T21:25:11Z")
> [@cderesende](#): > > initial\_estimation\_checks:: particle filtering requires measurement error on the observables Many thanks! For some reason, I missed that part and focused on the initial values for the estimation block.