# Problem with estimating on empirical data **URL:** \ **Category:** ML/Bayesian estimation\ **Created:** [8 November 2020 15:30 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899 "2020-11-08T15:30:34Z")\ **Posts on this page:** 20\ **Page:** 1
**Author:** ![helenaholm](https://forum.dynare.org/letter_avatar_proxy/v4/letter/h/e47774/32.png) [@helenaholm](https://forum.dynare.org/u/helenaholm)\ **Post date:** [8 November 2020 15:30 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/1 "2020-11-08T15:30:34Z")
Hi, I’m having some troubles with estimating a baseline version of the RBC-model i Dynare on empirical data (i.e. “baseline\_data” file). Can someone please help me why Dynare failure to calculate the estimated model? Thanks for taking your time. /Helena [romer\_estimation.mod](https://forum.dynare.org/uploads/short-url/do13XXCgxdg5k6Mm8n8nbh9FaFa.mod) (1.2 KB) [Baseline\_data.xlsx](https://forum.dynare.org/uploads/short-url/l3jflJ6HgjfNRmEZO0uEXuFUwSe.xlsx) (12.1 KB) ---
**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:** [8 November 2020 16:31 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/2 "2020-11-08T16:31:11Z")
Before moving to estimation, make sure you can simulate the model. For example, I am pretty sure the timing of capital in ```auto c = w*lab+(r+1-cdelta)*k - k(-1); ``` is wrong. ---
**Author:** ![helenaholm](https://forum.dynare.org/letter_avatar_proxy/v4/letter/h/e47774/32.png) [@helenaholm](https://forum.dynare.org/u/helenaholm)\ **Post date:** [8 November 2020 17:32 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/3 "2020-11-08T17:32:41Z")
I can simulate the model with no problems 🙂 ---
**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:** [8 November 2020 18:37 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/4 "2020-11-08T18:37:06Z")
How? `model_diagnostics` complains about `k` not being present at time `t`. ---
**Author:** ![helenaholm](https://forum.dynare.org/letter_avatar_proxy/v4/letter/h/e47774/32.png) [@helenaholm](https://forum.dynare.org/u/helenaholm)\ **Post date:** [8 November 2020 19:08 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/5 "2020-11-08T19:08:59Z")
The estimation is based on these simulations [rbc\_romer.mod](https://forum.dynare.org/uploads/short-url/vNvBg9kfOk74bEcI3tb8aV9CV9K.mod) (1.0 KB) ---
**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:** [8 November 2020 19:20 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/6 "2020-11-08T19:20:57Z")
How about [rbc\_romer.mod](https://forum.dynare.org/uploads/short-url/nIclIj9nJuLb2mrvwLUE7CiHm0o.mod) (1.0 KB) ? Your model has output going down after a positive TFP shock… ---
**Author:** ![helenaholm](https://forum.dynare.org/letter_avatar_proxy/v4/letter/h/e47774/32.png) [@helenaholm](https://forum.dynare.org/u/helenaholm)\ **Post date:** [8 November 2020 19:42 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/7 "2020-11-08T19:42:48Z")
Thanks for your input - the results changes especially the impulse response functions. Hence i still have implications with estimation on empirical data ![Skærmbillede 2020-11-08 kl. 20.40.37](https://forum.dynare.org/uploads/default/original/2X/e/e7a0cc1085bdf867222eceb90e754c624e40467f.png) ---
**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 November 2020 07:56 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/8 "2020-11-09T07:56:09Z")
Do you still have the `steady_state_model`-block in your mod-file? ---
**Author:** ![helenaholm](https://forum.dynare.org/letter_avatar_proxy/v4/letter/h/e47774/32.png) [@helenaholm](https://forum.dynare.org/u/helenaholm)\ **Post date:** [9 November 2020 08:12 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/9 "2020-11-09T08:12:13Z")
Above i estimated the model with the “steady” command and no “steady\_state\_model-blok”. Hence I tried again with the steady\_state\_model-blok where i obtain this _Error using print\_info (line 32)_ _The steadystate file did not compute the steady state_ [romer\_estimation.mod](https://forum.dynare.org/uploads/short-url/sJtLA3NLcUM36rDRXJRg6pqtlpt.mod) (1.6 KB) ---
**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 November 2020 08:21 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/10 "2020-11-09T08:21:39Z")
Use [romer\_estimation.mod](https://forum.dynare.org/uploads/short-url/jNQXaFRBRlCmQ9DxnS6SscDV3Eg.mod) (1.6 KB) `steady` will not work without proper initial values. Now your data treatment looks wrong. See [Pfeifer(2013): “A Guide to Specifying Observation Equations for the Estimation of DSGE Models”](https://drive.google.com/file/d/1r89OU5OE3CBa6tOlj6l3hNVWEaRH5Anv/view?usp=sharing) ---
**Author:** ![helenaholm](https://forum.dynare.org/letter_avatar_proxy/v4/letter/h/e47774/32.png) [@helenaholm](https://forum.dynare.org/u/helenaholm)\ **Post date:** [9 November 2020 09:25 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/11 "2020-11-09T09:25:53Z")
As I have 8 observable variables should I then declare 8 shocks? [romer\_estimation\_new.mod](https://forum.dynare.org/uploads/short-url/e8xu57OduPV3fkRLe0EbGvcoTHF.mod) (1.6 KB) ---
**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 November 2020 09:49 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/12 "2020-11-09T09:49:07Z")
If you really want 8 observables, you need at least 8 shocks. You may still get a problem with stochastic singularity if you don’t have some measurement error. All of that is detailed in the guide linked above. ---
**Author:** ![helenaholm](https://forum.dynare.org/letter_avatar_proxy/v4/letter/h/e47774/32.png) [@helenaholm](https://forum.dynare.org/u/helenaholm)\ **Post date:** [9 November 2020 11:46 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/13 "2020-11-09T11:46:27Z")
Yes that was also my thoughts - thank for providing me feedback, I really appreciate your help. I have now declared 2 observable variables and 2 shocks where some of the estimation process succeed but I estimate a negative matrix \<0 . “Error using chol Matrix must be positive definite.” How is it possible to correct for this? I tried to look at your guide but with no success. [romer\_estim\_old.mod](https://forum.dynare.org/uploads/short-url/cEM84XBx7LtKYwvsqGqfyXOLPwf.mod) (1.6 KB) ---
**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 November 2020 12:53 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/14 "2020-11-09T12:53:28Z")
The data file is missing. ---
**Author:** ![helenaholm](https://forum.dynare.org/letter_avatar_proxy/v4/letter/h/e47774/32.png) [@helenaholm](https://forum.dynare.org/u/helenaholm)\ **Post date:** [12 November 2020 08:38 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/15 "2020-11-12T08:38:31Z")
sorry for the many questions but i don’t understand how the data file can be missing when i have declared the data file in the estimation-block and the program are able to estimate part of the code? Thanks for taking your time to help me! [romer\_estim\_old.mod](https://forum.dynare.org/uploads/short-url/rgfmrEHILlCHHsSqpjfWooW4bhv.mod) (1.6 KB) ---
**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 November 2020 08:50 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/16 "2020-11-12T08:50:32Z")
What error message do you get? Note that programming languages are case-sensitive. Your Excel-file starts with a capital letter. ---
**Author:** ![helenaholm](https://forum.dynare.org/letter_avatar_proxy/v4/letter/h/e47774/32.png) [@helenaholm](https://forum.dynare.org/u/helenaholm)\ **Post date:** [12 November 2020 09:12 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/17 "2020-11-12T09:12:51Z")
_POSTERIOR KERNEL OPTIMIZATION PROBLEM!_ - (minus) the hessian matrix at the “mode” is not positive definite!\* _=\> posterior variance of the estimated parameters are not positive._ _You should try to change the initial values of the parameters using_ _the estimated\_params\_init block, or use another optimization routine._ \*Warning: The results below are most likely wrong! \* _\> In dynare\_estimation\_1 (line 315)_ - In dynare\_estimation (line 105)\* - In romer\_estim\_old.driver (line 247)\* - In dynare (line 293) \* _MODE CHECK_ _Fval obtained by the minimization routine (minus the posterior/likelihood)): 13.409097_ _Warning: Matrix is singular, close to singular or badly scaled. Results may be_ \*inaccurate. RCOND = NaN. \* _\> In dynare\_estimation\_1 (line 338)_ - In dynare\_estimation (line 105)\* - In romer\_estim\_old.driver (line 247)\* - In dynare (line 293)\* _Log data density [Laplace approximation] is NaN._ _Error using chol_ _Matrix must be positive definite._ _Error in posterior\_sampler\_initialization (line 84)_ _d = chol(vv);_ _Error in posterior\_sampler (line 60)_ - posterior\_sampler\_initialization(TargetFun, xparam1, vv,\* - mh\_bounds,dataset\_,dataset\_info,options\_,M\_,estim\_params\_,bayestopt\_,oo\_);\* _Error in dynare\_estimation\_1 (line 465)_ - ``` posterior_sampler(objective_function,posterior_sampler_options.proposal_distribution,xparam1,posterior_sampler_options,bounds,dataset_,dataset_info,options_,M_,estim_params_,bayestopt_,oo_);* ``` _Error in dynare\_estimation (line 105)_ - dynare\_estimation\_1(var\_list,dname);\* _Error in romer\_estim\_old.driver (line 247)_ _oo\_recursive\_=dynare\_estimation(var\_list\_);_ _Error in dynare (line 293)_ _evalin(‘base’,[fname ‘.driver’]) ;_ I’m trying to estimate the model with more than one observable variables (y\_obs and c\_obs) - when i estimate with only y\_obs i receive no error messages so i think it could be in the definition of the additional shock variable/equation where it goes wrong. 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:** [12 November 2020 09:50 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/18 "2020-11-12T09:50:16Z")
Please provide the most recent mod-file and data-file ---
**Author:** ![helenaholm](https://forum.dynare.org/letter_avatar_proxy/v4/letter/h/e47774/32.png) [@helenaholm](https://forum.dynare.org/u/helenaholm)\ **Post date:** [12 November 2020 10:20 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/19 "2020-11-12T10:20:54Z")
[romer\_estim\_old.mod](https://forum.dynare.org/uploads/short-url/s5unVz8fV9j6APgyx1J3lAnH2R0.mod) (1.6 KB) [baseline\_romer.m](https://forum.dynare.org/uploads/short-url/uKdxH0JVuz9LFMOSdTfo9VwVent.m) (1.4 KB) 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:** [12 November 2020 11:00 UTC](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899/20 "2020-11-12T11:00:48Z")
I am not sure what you are doing. 1. The uploaded file does not run, because there is a log missing in ```auto log(cA) = rhov*(cA(-1))+epsv; ``` 1. The parameter `rhov` is clearly not identified as you are neither observing `cA` nor is `cA` affecting any other part of the model. [Next page](https://forum.dynare.org/t/problem-with-estimating-on-empirical-data/16899.md?page=2)