a question of a diffuse prior

[kenandaomi]

Dear Junior, I encountered a problem in model estimation, and the results show like this. I read the previous answer about the estimation and revised it, but it doesn't seem to work. I have used different solutions, including the ergodic mean and a particular point, but I always get the following prompt: Exploration limit reached. I tried the method in # 13, 'kf init Variance ', 5 and' kf_ Presample ', arbitraryInteger, but no matter where I put it, it always doesn't work. If you have time to look at the code for me, I will be very grateful.

01.zip

[jmaih]

Hi,

There is at least one major problem: You cannot use 3 observable variables (Y, r, pp) on a model that has only 2 shocks (e_a, e_z). This inevitably creates a problem of stochastic singularity.

Cheers,

J.

On Sun, Mar 5, 2023 at 1:09 PM kenandaomi @.***> wrote:

Dear Junior, I encountered a problem in model estimation, and the results show like this. [image: 报错] https://user-images.githubusercontent.com/105027866/222959384-cc14bbb9-8045-462c-9330-c6218107e0bd.png I read the previous answer about the estimation and revised it, but it doesn't seem to work. I have used different solutions, including the ergodic mean and a particular point, but I always get the following prompt: Exploration limit reached. I tried the method in # 13, 'kf init Variance ', 5 and' kf_ Presample ', arbitraryInteger, but no matter where I put it, it always doesn't work. If you have time to look at the code for me, I will be very grateful.

01.zip https://github.com/jmaih/RISE_toolbox/files/10891778/01.zip

— Reply to this email directly, view it on GitHub https://github.com/jmaih/RISE_toolbox/issues/178, or unsubscribe https://github.com/notifications/unsubscribe-auth/AATKBTYCXANLCRGXS3LU2A3W2R7A5ANCNFSM6AAAAAAVQEDTMI . You are receiving this because you are subscribed to this thread.Message ID: @.***>

[kenandaomi]

Thank you for your quick reply, but in fact, I only used r as an observation variable in the simulation, and still generated the error "Exploration limit reached" when I estimated. I don't know much about the estimation process, which may cause many problems in the m file "run01_solution". If possible, could you please give some suggestions about this file. Thanks again

[jmaih]

Hi,

I don't understand when you say " Thank you for your quick reply, but in fact, I only used r as an observation variable in the simulation"

Estimation is not the same as simulation. In the model file, you declared three observable variables.

If, as you said, you don't understand the estimation process, I would recommend you try and understand what the examples do first. Even still, before you get to estimation, you should be able to at least solve the model first.

Cheers,

J.

On Mon, Mar 6, 2023 at 3:31 PM kenandaomi @.***> wrote:

Thank you for your quick reply, but in fact, I only used r as an observation variable in the simulation, and still generated the error "Exploration limit reached" when I estimated. I don't know much about the estimation process, which may cause many problems in the m file "run01_solution". If possible, could you please give some suggestions about this file. Thanks again

— Reply to this email directly, view it on GitHub https://github.com/jmaih/RISE_toolbox/issues/178#issuecomment-1456245548, or unsubscribe https://github.com/notifications/unsubscribe-auth/AATKBT3EW2WKZ25CM5RWRNLW2XYL5ANCNFSM6AAAAAAVQEDTMI . You are receiving this because you commented.Message ID: @.***>

[kenandaomi]

Hi，

I relearned the examples in the toolbox and understood the difference between simulation and estimation. I have readjusted my code, but unfortunately, it still did not succeed. If you have time to help me take a look at my code, I would be very grateful. By the way, I don't seem to have found any examples of using Bayesian estimation in the toolbox. I tried the code related to Bayesian estimation in Dynare, but RISE seems to be unable to recognize it. How do I need to adjust it? Thanks again

------------------ 原始邮件 ------------------ 发件人: "jmaih/RISE_toolbox" @.>; 发送时间: 2023年3月6日(星期一) 晚上10:41 @.>; @.**@.>; 主题: Re: [jmaih/RISE_toolbox] a question of a diffuse prior (Issue #178)

Hi,

I don't understand when you say " Thank you for your quick reply, but in fact, I only used r as an observation variable in the simulation"

Estimation is not the same as simulation. In the model file, you declared three observable variables.

If, as you said, you don't understand the estimation process, I would recommend you try and understand what the examples do first. Even still, before you get to estimation, you should be able to at least solve the model first.

Cheers,

J.

On Mon, Mar 6, 2023 at 3:31 PM kenandaomi @.***> wrote:

> Thank you for your quick reply, but in fact, I only used r as an > observation variable in the simulation, and still generated the error > "Exploration limit reached" when I estimated. I don't know much about the > estimation process, which may cause many problems in the m file > "run01_solution". If possible, could you please give some suggestions about > this file. > Thanks again > > — > Reply to this email directly, view it on GitHub > <https://github.com/jmaih/RISE_toolbox/issues/178#issuecomment-1456245548&gt;, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AATKBT3EW2WKZ25CM5RWRNLW2XYL5ANCNFSM6AAAAAAVQEDTMI&gt; > . > You are receiving this because you commented.Message ID: > @.***> >

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

[jmaih]

I think I can close this issue. Feel free to reopen it if you still have problems