# Presample

**URL:** <https://forum.dynare.org/t/presample/90>\
**Category:** Dynare help (legacy posts)\
**Created:** [8 May 2006 20:16 UTC](https://forum.dynare.org/t/presample/90 "2006-05-08T20:16:16Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![jmaih](https://forum.dynare.org/letter_avatar_proxy/v4/letter/j/3d9bf3/32.png) [@jmaih](https://forum.dynare.org/u/jmaih)\
**Post date:** [8 May 2006 20:16 UTC](https://forum.dynare.org/t/presample/90/1 "2006-05-08T20:16:16Z")

</div>

Hi,  
I have a version of dynare 3 that computes the filtered and smoothed variables even for the presample period. Is that the way things should be?

Junior

---

<div class="post-metadata">

**Author:** ![MichelJuillard](https://forum.dynare.org/user_avatar/forum.dynare.org/micheljuillard/32/5318_2.png) [@MichelJuillard](https://forum.dynare.org/u/MichelJuillard)\
**Post date:** [9 May 2006 11:43 UTC](https://forum.dynare.org/t/presample/90/2 "2006-05-09T11:43:34Z")

</div>

Yes, I don’t see a problem there. The ‘presample’ periods are periods at the beginning of the sample, after initialization of the filter but that aren’t used to compute the loglikelihood. Filtered values must be computed for them and I don’t see why not computing smoothed values as well.

Kind regards

Michel

---

<div class="post-metadata">

**Author:** ![jmaih](https://forum.dynare.org/letter_avatar_proxy/v4/letter/j/3d9bf3/32.png) [@jmaih](https://forum.dynare.org/u/jmaih)\
**Post date:** [9 May 2006 13:01 UTC](https://forum.dynare.org/t/presample/90/3 "2006-05-09T13:01:08Z")

</div>

Dear Michel,  
Thanks a lot.

Junior