# IRF matching with gaps

**URL:** <https://forum.dynare.org/t/irf-matching-with-gaps/28195>\
**Category:** SMM/GMM/IRF-Matching\
**Created:** [20 January 2026 17:15 UTC](https://forum.dynare.org/t/irf-matching-with-gaps/28195 "2026-01-20T17:15:51Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![canderl](https://forum.dynare.org/letter_avatar_proxy/v4/letter/c/bbce88/32.png) [@canderl](https://forum.dynare.org/u/canderl)\
**Post date:** [20 January 2026 17:15 UTC](https://forum.dynare.org/t/irf-matching-with-gaps/28195/1 "2026-01-20T17:15:51Z")

</div>

Hello,  
I would like to know whether it is possible to match against an IRF which has gaps? For example in one of our use cases we have values for an empirical IRF for horizons 1 and 2 then nothing for horizon 3 and 4 and then again values for horizons 5 and 6. Is there functionality to allow matching to such an IRF with gaps?

---

<div class="post-metadata">

**Author:** ![Brasidas](https://forum.dynare.org/letter_avatar_proxy/v4/letter/b/d2c977/32.png) [@Brasidas](https://forum.dynare.org/u/Brasidas)\
**Post date:** [21 January 2026 09:19 UTC](https://forum.dynare.org/t/irf-matching-with-gaps/28195/2 "2026-01-21T09:19:42Z")

</div>

One way of dealing with this is to adjust the weighting matrix, attaching a weight of zero to the IRFs in periods 3 and 4 whilst still inputting some values to ensure that the length of the IRF vector is consistent with that of the model.

---

<div class="post-metadata">

**Author:** ![canderl](https://forum.dynare.org/letter_avatar_proxy/v4/letter/c/bbce88/32.png) [@canderl](https://forum.dynare.org/u/canderl)\
**Post date:** [22 January 2026 10:38 UTC](https://forum.dynare.org/t/irf-matching-with-gaps/28195/3 "2026-01-22T10:38:53Z")

</div>

Great advice - this has solved my problem. Thank you very much!