Mode check plots and mode

Thank you Prof. Pfeifer, for your prompt reply. I really appreciate it.

Dear All,

I am estimating my parameters using in the Method of moments toolbox whereby the mode_check displays log-dist kernel. I wonder if it is usual to have a plot or two with kink. How am I suppose to interpret them? Do they indicate that my model has problems?

Thank you in advance!

The kinks are indeed rather unusual but arise from the restricted domain on the x axis. You need to find out why you have so many red dots and what causes them. Looking at them, you may have forgotten to restrict the domain to [0,1] for the autocorrelation parameters.

1 Like

Thank sm you for the reply. Indeed, I didn’t restrict the domain for the autocorrelation parameters. Thank you for pointing this out!