Puzzle with the Log-linearized

In DYNARE, if we write the variables into exp(.) form, then why we need Log-linearized model by hand. By the way, I’m a new beginner with the DYNARE.

Sorry, but I don’t understand your question. But with Dynare, there is typically no need for manual linearization. See also Question about understanding irfs in dynare

Professor Jpfeifer, as you say there is typically no need for manual linearization in DYNARE. But I don’t understand why some people write Log-linearized models in Dynare. For simplify, we can write the original equations such as FOC into dynare directly. Thanks a lot , my best.

Some people do it, because they grew up having to do the linearization manually. But the most important reason is that there are problems that sometimes do not have a closed-form, finite nonlinear representation. See e.g. How to write a infinite sum FOC into dynare

Thank you so much, I get it.
Best wishes,