# Question on convergence diagnostics

**URL:** <https://forum.dynare.org/t/question-on-convergence-diagnostics/19460>\
**Category:** Estimation\
**Created:** [12 December 2021 09:58 UTC](https://forum.dynare.org/t/question-on-convergence-diagnostics/19460 "2021-12-12T09:58:50Z")\
**Posts on this page:** 8\
**Page:** 1

<div class="post-metadata">

**Author:** ![louisvalentines](https://forum.dynare.org/letter_avatar_proxy/v4/letter/l/dec6dc/32.png) [@louisvalentines](https://forum.dynare.org/u/louisvalentines)\
**Post date:** [12 December 2021 09:58 UTC](https://forum.dynare.org/t/question-on-convergence-diagnostics/19460/1 "2021-12-12T09:58:50Z")

</div>

Dear professors, I estimated a DSGE model in dynare, then the dynare plotted the convergence diagnostic . And my question is that convergence diagnostic is necessary to judge the confidence of the estimation results ? If multivariate convergence diagnostic seems not convergence but the acceptance rate is in 25%-33% , should I believe the estimation results?

---

<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:** [12 December 2021 16:28 UTC](https://forum.dynare.org/t/question-on-convergence-diagnostics/19460/2 "2021-12-12T16:28:11Z")

</div>

No, if the chain has not yet converged to the ergodic distribution, the results are meaningless, regardless of the acceptance rate.

---

<div class="post-metadata">

**Author:** ![louisvalentines](https://forum.dynare.org/letter_avatar_proxy/v4/letter/l/dec6dc/32.png) [@louisvalentines](https://forum.dynare.org/u/louisvalentines)\
**Post date:** [12 December 2021 16:52 UTC](https://forum.dynare.org/t/question-on-convergence-diagnostics/19460/3 "2021-12-12T16:52:17Z")

</div>

Thanks professor for your reply. The convergence of chain seems like playing tricks, is there any good way to get reliable results?

---

<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:** [12 December 2021 18:20 UTC](https://forum.dynare.org/t/question-on-convergence-diagnostics/19460/4 "2021-12-12T18:20:18Z")

</div>

What do the `trace_plot` say? Is there a drift in the chain?

---

<div class="post-metadata">

**Author:** ![louisvalentines](https://forum.dynare.org/letter_avatar_proxy/v4/letter/l/dec6dc/32.png) [@louisvalentines](https://forum.dynare.org/u/louisvalentines)\
**Post date:** [13 December 2021 03:34 UTC](https://forum.dynare.org/t/question-on-convergence-diagnostics/19460/5 "2021-12-13T03:34:43Z")

</div>

Here are my mod file and data file, and trace plot.  
The data I used, are took log then used 1st HP filter detrend.  
[logtest.mod](https://forum.dynare.org/uploads/short-url/w8oUMA2Hbwe3RfLDFy2FX3OP5Xh.mod) (3.7 KB)  
[gdata.m](https://forum.dynare.org/uploads/short-url/rwUGiTMJvF5mgZUfB6V30oHWzBT.m) (7.3 KB)  
[block number 1.eps](https://forum.dynare.org/uploads/short-url/vG9bzK0FLOzwWqNOO2FwHCEnxY8.eps) (700.7 KB)  
[block number 2.eps](https://forum.dynare.org/uploads/short-url/w7MTKdXAt8cZIp2lMJ5Sp5cnLk8.eps) (699.1 KB)  
[logtest\_mdiag.eps](https://forum.dynare.org/uploads/short-url/dEYwpw68AR1ewK7tDxfqXGqBhac.eps) (32.8 KB)

[pictures.zip](https://forum.dynare.org/uploads/short-url/cfbMaQqvHFaVqdax8fDzJxXXKdw.zip) (620.7 KB)

---

<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:** [13 December 2021 07:58 UTC](https://forum.dynare.org/t/question-on-convergence-diagnostics/19460/6 "2021-12-13T07:58:17Z")

</div>

This mostly looks like you may need a lot more draws.

---

<div class="post-metadata">

**Author:** ![louisvalentines](https://forum.dynare.org/letter_avatar_proxy/v4/letter/l/dec6dc/32.png) [@louisvalentines](https://forum.dynare.org/u/louisvalentines)\
**Post date:** [13 December 2021 08:51 UTC](https://forum.dynare.org/t/question-on-convergence-diagnostics/19460/7 "2021-12-13T08:51:32Z")

</div>

I increased the number of draws to 200000. It seems still not converge.  
Here are plots.  
[mucodiag.eps](https://forum.dynare.org/uploads/short-url/pQ36bTUNTMlemT5A69KJthTm5fp.eps) (33.1 KB)  
[block 2.eps](https://forum.dynare.org/uploads/short-url/cpnCQ71CFCJZQrGtp77uDZPxxi4.eps) (278.1 KB)  
[block 1.eps](https://forum.dynare.org/uploads/short-url/rGJuxlEm5oGp6q9A2lWx3UPkhcN.eps) (278.2 KB)

---

<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:** [13 December 2021 08:59 UTC](https://forum.dynare.org/t/question-on-convergence-diagnostics/19460/8 "2021-12-13T08:59:15Z")

</div>

Now it doesn’t look too bad. I would run chain with 1 million draws.