lgarrison

lgarrison

lgarrison

#!/usr/bin/env python3
'''
Compress a density grid.

Usage: ./compress.py --help
'''

from pathlib import Path

import click

lgarrison

lgarrison

lgarrison

import numpy as np
import threading
import queue

def multithreaded_read(fns, nthreads=2):
    '''
    A simple multithreaded reader that spins up N threads
    to read a list of files, returning the result in a list.
    Reads in Numpy format by default.
    

lgarrison

import numba as nb
import numpy as np

def nb_ecdf(X,Rmax,ngrid=64,nthreads=1):
    '''Compute the empirical CDF of points X (shape (N,D))
    in domain [0,Rmax) using a histogram with `ngrid` cells
    per dimension.
    
    Rmax and ngrid can also be length D arrays.
    '''

lgarrison

'''
A simple multi-threaded Python file writer.

Author: Lehman Garrison (lgarrison.github.io)
'''

import threading
import queue

import numpy as np

lgarrison

#!/usr/bin/env python3

'''
This script facilitates Globus transfers of Abacus simulation data.
In particular, for AbacusSummit we want to transfer particle data and
halo catalogs from OLCF to NERSC after each simulation is done.
Both sites already have Globus endpoints, so the job of this script
is to initiate a transfer between the two, saving the transfer ID
number so we can query the transfer status later.
The intention of this script is primarily to do disk-to-disk

lgarrison

/*

This is a "minimal" reproducer for an apparent GCC-7 bug related to AVX-512.
It has a 100% reproduction rate on my machine (Cascade Lake).

A simple loop that copies the values from a double array to a float array
sometimes gives the wrong answer in the first few elements (zero, nan, or garbage).

The bug seems correlated with alignment values of the input and output
arrays, but it's not a one-to-one relationship.