michael-petersen

Benchmarking EXP

For use with computing time proposals

michael-petersen

Brownian Noise

identification in the wild

I spend a lot of time sitting in these sound isolating phone booths. And one of the early things you might notice while sitting in them is that there is quite a bit of noise. And let's be honest, I don't pay attention during all of the calls, so when my mind wandered, it drifted towards some projects I've been working on...

I've been wondering about whether the booths are purposely trying to get me to focus. There were quite a few articles this past autumn about broadband noise with the goal of focusing your mind. In particular, brownian (or red) noise as all the rage, see WaPo, NYT, Guardian coverage. And when I read the articles, I wondered: is the booth pumping focusing

michael-petersen

Making GNU and Apple play nice

Congratulations! You have now committed to turning your arm64 machine into a GNU development device. This is not trivial, but we seem to have solved the bulk of the issues. These notes are mainly relevant for installing EXP, although how to install Gadget-4 basically happens for free.

As of right now (2-May-2023), you must install the main branch of EXP in order to compile on arm64 machines.

A few guiding points:

• Don't forget to git submodule update --init --recursive!
• If you think it's a bug, it more likely is a path issue.
• clang is pretty finicky with warnings; most are no cause for panic.

michael-petersen

# standard python modules
import numpy as np

# plotting utilities
import matplotlib.pyplot as plt
     
def trace_points(x,y,startpoint=0):
    """trace_points
    starting from a defined point in a set, find the next closest point, to make a continous line
    """

michael-petersen

import matplotlib.pyplot as plt

# the code in rotation.py
from rotation import *

# the code in flatdisc.py
from flatdisc import *

N = 10000
x,y,z,u,v,w = generate_basic_disc_points(N)

michael-petersen

"""
This version uses constants:
  ugly from an object-oriented standpoint, but fast.
"""

using BenchmarkTools
"""
Follow along
https://docs.juliahub.com/BenchmarkTools/ZXPQo/1.0.0/
"""

michael-petersen

Galaxy,Morph.Type,z,M_r,Rbar,Rbarh,Rbarl,Sbar,Omega,eOmega,Rcr,Rcrh,Rcrl,R,Rl,Rh,B/T(Ref.),Ref.,FinalSample,Rnew,Rnewl,Rnewh
ESO 139-G09 , SAB0^0(rs) ,0.018,-21.2,6.41,2.41,-1.13,0.44,56.7,15.4,5.54,1.96,-1.17,0.8,0.3,-0.2, ... ,1, yes,,,
ESO 281-G31 , SB0^0(rs) ,0.018,-21.22,4.04,0.37,-0.37,,28.6,11.2,7.35,4.41,-1.47,1.8,1.1,-0.4, ... ,2, yes,,,
IC 874 , SB0^0(rs) ,0.008,-20.57,3.65,0.94,-0.9,0.57,38.2,13.1,4.93,2.4,-1.25,1.4,0.7,-0.4, ... ,1, yes,,,
NGC 271 , (R’)SBab(rs) ,0.014,-21.81,7.7,0.27,-0.27,0.63,29.5,16,11.69,7.97,-4.25,1.5,1,-0.5, ... ,2, no,,,
NGC 936 , SB0^+(rs) ,0.005,-20.71,4.16,,,0.55,56.8,13.2,5.74,1.25,-1.25,1.4,0.5,-0.4, 0.20 (1) ,3, yes,,,
NGC 1023 , SB0^-(rs) ,0.002,-21.39,3.94,0.29,-0.29,0.48,89.2,31.5,3.03,0.34,-0.34,0.8,0.5,-0.3, ... ,4, yes,,,
NGC 1308 , SB0/a(r) ,0.021,-21.71,5.36,0.78,-1.47,0.5,91.8,32.1,3.85,1.95,-0.99,0.8,0.4,-0.2, ... ,1, yes,,,
NGC 1358 , SAB0/a(r) ,0.013,-21.09,5.16,0.81,-0.81,,34.4,16.6,6.24,5.16,-1.9,1.2,1,-0.4, ... ,2,  yes,,,
NGC 1440 , (R')SB0^0 ,0.005

michael-petersen

#!/bin/sh

# we want to create a grid of simulations, from
# lmax \in [2,4,6,8,12,16,24,32,64]
# nmax \in [2,4,8,16,24,32,48,64,128]

# for an ensemble of 10 simulations, e.g.
# run in run1a run1b run1c run1d run1e run1f run1g run1h run1i run1j

for run in run2a run2b run2c run2d run2e run2f

michael-petersen

import numpy as np

def gen_sphere_surface(npoints):
    """generate a sphere of stars, evenly spaced on the surface of a sphere"""
    npoint = 0
    points = np.zeros([npoints,3])
    a = 4*np.pi/npoints
    d = np.sqrt(a)
    Mtheta = np.floor(np.pi/d).astype('int')
    dtheta = np.pi/Mtheta

michael-petersen

import numpy as np

def return_density(logr,weights=1.,rangevals=[-2, 6],bins=500,d2=False):
    """return_density
    
    simple binned density using logarithmically spaced bins
    
    inputs
    ---------
    logr        : (array) log radii of particles to bin