# Subsamples estimation

**URL:** <https://forum.dynare.org/t/subsamples-estimation/5913>\
**Category:** Dynare help (legacy posts)\
**Created:** [5 December 2016 16:03 UTC](https://forum.dynare.org/t/subsamples-estimation/5913 "2016-12-05T16:03:39Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![irish](https://forum.dynare.org/letter_avatar_proxy/v4/letter/i/eada6e/32.png) [@irish](https://forum.dynare.org/u/irish)\
**Post date:** [5 December 2016 16:03 UTC](https://forum.dynare.org/t/subsamples-estimation/5913/1 "2016-12-05T16:03:39Z")

</div>

Dear all,

I would like to estimate a model which exhibits two different growth trends over the observation period. I have read a lot of posts in this forum and many related suggested readings but still can not figure out how to manage with this. I need my model to grow at rate, say, of 1% before 1980 and of 2% thereafter. My topic may look stupid but I would really appreciate any help.  
regards,  
irish

---

<div class="post-metadata">

**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:** [5 December 2016 18:31 UTC](https://forum.dynare.org/t/subsamples-estimation/5913/2 "2016-12-05T18:31:40Z")

</div>

You need to provide more details. Do you want to estimate parameters related to the structural break? In general, this is rather involved and still on our to-do list. See [dynare.org/DynareWiki/SubsamplesEstimation](http://www.dynare.org/DynareWiki/SubsamplesEstimation)

---

<div class="post-metadata">

**Author:** ![irish](https://forum.dynare.org/letter_avatar_proxy/v4/letter/i/eada6e/32.png) [@irish](https://forum.dynare.org/u/irish)\
**Post date:** [6 December 2016 11:51 UTC](https://forum.dynare.org/t/subsamples-estimation/5913/3 "2016-12-06T11:51:28Z")

</div>

Dear mr. Johannes,

thank you very much for your kind reply. I need to estimate the model parameters and to compute the log-posterior over the whole sample period. My log-lin model is characterized by a composite growth trend, i.e. a linear combination of two exhogenous stochastic processes. In steady state, the the growth rates of these two shocks take on different values before and after a certain year. So, I would like to estimate these four different trend values together with the other model parameters, which maintaint the intitialized calibration over the whole sample. You can kindly find the mod file attached. It seems to run the estimation, but I am not sure the coding is correct. I would really appreaciate your feedback on it.  
thanks a lot.

---

<div class="post-metadata">

**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:** [9 December 2016 08:41 UTC](https://forum.dynare.org/t/subsamples-estimation/5913/4 "2016-12-09T08:41:02Z")

</div>

No, what you do does not work. You currently cannot do structural breaks easily without seriously changing some of Dynare’s internal codes. I will try to work on this in the next weeks, but it will take some time.

---

<div class="post-metadata">

**Author:** ![irish](https://forum.dynare.org/letter_avatar_proxy/v4/letter/i/eada6e/32.png) [@irish](https://forum.dynare.org/u/irish)\
**Post date:** [12 December 2016 15:39 UTC](https://forum.dynare.org/t/subsamples-estimation/5913/5 "2016-12-12T15:39:27Z")

</div>

Dear mr. Johannes,

thank you very much for your kind reply and availability.  
Looking forward to hearing from you.  
Best regards,  
irish