Use only min mass gas in time step

Fix #758 (closed)

My student finished its simulations and everything seems alright. Do you wish to keep the current behavior for your simulations or are you fine if merge it like this?

added GEAR SPH performance labels

Please clarify what "everything seems alright" means.

At a minimum, I want to see a baryon matter power-spectrum with and without these changes from a representative volume.

The issue was about a huge slow down. Now the simulations are taking the same amount of time with and without Pop III stars.

I agree that this is an improvement. But this code change may have consequences for everybody else using the code. Including my own simulations. I want to know that the main things I care about are not affected.

I will see if we can get a nice power spectrum from our simulations.

Would you prefer a parameter that allows you to switch between the minimal mass of all the baryons and only from the gas rather than forcing the change to everyone?

What is the status of this MR?

I used to have a student that was working on it, but I have no idea what he is doing right now. @yvesrevaz Do you know what is going on?

This issue was not a priority for him. So, nothing has been done. I think to move on on this point, the best would be to have a parameter that allow to switch, as suggested by Loic, without breaking anything. Obtaining a power spectrum would be of course ideal, but I have unfortunately no manpower right know to do this test, and at Lastro, we have no simple way to extract the power spectrum from a Swift output. If someone from Swift has a tool fully ready to do it, I will be happy to try.

added 269 commits

• e93c9d61...068064ac - 267 commits from branch master
• e00d697e - Use only min mass gas in time step
• f378c456 - Add comment

Compare with previous version

added 3 commits

• db44b166 - Add a runtime parameter to control the behaviour.
• 0007a908 - Use the new parameter in the Agora GEAR examples
• e1c79f37 - Update the RTD for the new values

Compare with previous version

Ok. I have settled for a runtime parameter switching the behaviour from the default to the case where only the minimal gas mass is considered for the mesh RMS time-step condition. The parameter is switched on in the GEAR zoom example. If unspecified, it defaults to 0, i.e. use the old behaviour.

merged

mentioned in commit 8b8250fa

Perfect, thanks for taking care of this.