On Jul 29, 2008, at 3:50 AM, Jefferson Thomas wrote: > Thank you for your reply. > What suggestion do you mean ? I dont see any suggestion in Steven's > mail. > Sure I can calculate Q using longer intervals and it is exactly what I > am doing now.
Computing the energy in the entire cell and plotting it as a function of time is not a very good way to compute Q, in my experience. You really only need the field at a single point in the cavity as a function of time. And rather than fitting it yourself, it is usually better to use harminv (i.e. in my experience harminv will give you a more accurate Q with fewer data points than fitting the energy or field amplitude to a decaying exponential, especially if you use a relatively narrow-bandwidth source around your cavity frequency). (There are also other ways besides harminv, e.g. there is a nice method involving Pade approximants that is sometimes even better. But again, that only involves the fields at a single point in the cavity as a function of time.) > Of course I understand when you say that the generality has higher > priority. I perfectly agree. But I also think that finding > the total energy is a pretty common and important task and it deserves > its own optimized function Finding the total energy in the computation at *every time step* is not a common task in my experience. I personally have never needed to do so except in artificial testing situations. Doing it usually indicates that you are making a mistake, as seems to be the case here. Optimizing mistakes is definitely not a priority. Steven _______________________________________________ meep-discuss mailing list meep-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
