Snapshot Time Resolution and Smoothing Kernel for Gas Profile
Arnatri Samajdar
19 May
Hello I wanted to know where will I get the information regarding for how long each snapshot of TNG50-1 and TNG100-1 has been run? I want to get the time resolution to explain some of the results I am getting regarding effect of feedback in gas. Also I am studying the effect of feedback by categorizing the halo samples in 3 groups based on eddington ratio and bolometric luminosity and creating a stacked profile of each. Do you think I need to assume some kernel to do smoothing and then calculate weighted average of quantities (temperature, entropy etc) at each bin? or i can just do mass weighted average?
Dylan Nelson
19 May
The "snapshot spacing" i.e. the time (in Gyr, or redshift) between snapshots can be obtained by loading the header (metadata) of all the snapshots. There you find the Redshift and Time (scalefactor), so you can save them all in an array, then calculate the np.diff(), for example.
Note that "snapshot spacing" is independent from "time resolution" of the actual simulation.
For radial profiles, we rarely use spatial smoothing/kernels. Instead it is ok to simply bin distances, treating each gas cell as a point.
Whether you want to mass-weight, volume-weight, or otherwise weight, for mean profiles, depends on the science question to be addressed.
Hello I wanted to know where will I get the information regarding for how long each snapshot of TNG50-1 and TNG100-1 has been run? I want to get the time resolution to explain some of the results I am getting regarding effect of feedback in gas. Also I am studying the effect of feedback by categorizing the halo samples in 3 groups based on eddington ratio and bolometric luminosity and creating a stacked profile of each. Do you think I need to assume some kernel to do smoothing and then calculate weighted average of quantities (temperature, entropy etc) at each bin? or i can just do mass weighted average?
The "snapshot spacing" i.e. the time (in Gyr, or redshift) between snapshots can be obtained by loading the header (metadata) of all the snapshots. There you find the Redshift and Time (scalefactor), so you can save them all in an array, then calculate the
np.diff(), for example.Note that "snapshot spacing" is independent from "time resolution" of the actual simulation.
For radial profiles, we rarely use spatial smoothing/kernels. Instead it is ok to simply bin distances, treating each gas cell as a point.
Whether you want to mass-weight, volume-weight, or otherwise weight, for mean profiles, depends on the science question to be addressed.