Hi Sasan,
Thanks a lot for your suggestions last time!
Well, actually I'm looking at the nearshore wave breaking and propagation process in some areas of interest, and feel Celeris did better than SCHISM, the model we used before, which can't calculate the wave generated by plunging waves. However, to balance the resolution and timestep when input a long wave, it took much time to run (time ratio = 0.02). So... I'm curious that if it's possible to run without visualization? (It might not be a reasonable question, since Celeris is designed for visualization. ^v^|||)
Sincerely,
Chen
Unfortunately, that's right. It is not possible to restrict logging to a specific time.
Thanks Sasan! And hope to confirm, so, for now it is also impossible to log data from a specific time point, I suppose? (Since I'm looking at the final steady state of keeping one wave input, the data of first thousands of seconds are not necessary to be recorded.
Happy to hear that you are getting better results with Celeris. As you mentioned, Celeris is designed for visualization, so running it without visualization is not possible at this point. Note that visualization is not as computationally heavy as the actual numerical stuff. Try minimizing the window, that will effectively disable the visualization, at least most of it. Furthermore, if you turn on the logging, Celeris will run much slower. If you are logging data, try to log as minimally as possible. That means less frequently (larger log step) and sparsely (few gauges instead of large ranges).
Buying a more powerful GPU will also help :)