Help with non-linear estimation

Dear Prof.,

I am trying to run a non-linear estimation using the particle filter method. My code runs when simulated and model diagnostic reveals that there is no detected issue with the model. However, using both Dynare 4.6.4 and the unstable version, I keep getting the error message ‘I am not able to read the dates’

I already checked through related posts on this forum and attempted the suggestions but the code still does not run. Could you please help check through the attached .mod file for the possible remedy?
toy.mod (6.8 KB)
fpmpq1.xlsx (13.6 KB)

Cell A9 of the Excel sheet contains text instead of a number due to using a dot instead of a proper minus.

Thank you Prof for the useful suggestion. It solved the error.

However, I encountered another error message Error: chol: input matrix must be positive definite but I (1) adjusted the nonlinear_filter_initialization from 2 to 3 and (2) removed options_.particle.pruning=1, and the model seems to be running.

Am I doing this correctly?

Yes, but you need to decide whether to leave out the pruning.

Thank you Prof. Your suggestion is well received. I will compare the outputs under two scenarios: with pruning and without pruning and then decide.

Thank you and Best Regards.

Dear Prof.,

I want to clarify the message I get on Octave after running the code. It says:

‘Subsampling on first observation’. Is this a bug? or the code running?

I have never seen this message. Could you provide the full message and the file(s) that cause it.

Thank you Prof., this the message I get:

MODEL_DIAGNOSTICS: No obvious problems with this mod-file were detected.

Estimation using a non linear filter!

warning: initial_estimation_checks:: the pruning settings differ between the particle filter a
nd the one used for IRFs/simulations. Make sure this is intended.\n
warning: called from
initial_estimation_checks at line 75 column 9
dynare_estimation_1 at line 159 column 9
dynare_estimation at line 118 column 5
driver at line 736 column 14
dynare at line 281 column 5

Subsample with only the first observation.

toy.mod (6.8 KB)
fpmpq1.xlsx (13.5 KB)

Thank you Prof., after running my code on non-linear estimation.
I have also attached the code and datafile

This is the message I get:

MODEL_DIAGNOSTICS: No obvious problems with this mod-file were detected.

Estimation using a non linear filter!

warning: initial_estimation_checks:: the pruning settings differ between the particle filter a
nd the one used for IRFs/simulations. Make sure this is intended.\n
warning: called from
initial_estimation_checks at line 75 column 9
dynare_estimation_1 at line 159 column 9
dynare_estimation at line 118 column 5
driver at line 736 column 14
dynare at line 281 column 5

Subsample with only the first observation.

toy.mod (6.8 KB)
fpmpq1.xlsx (13.5 KB)

Yes, it is running, but enters an infinite loop because your steady state file does not work for all parameter values.