0x0L

import numpy as np

N = 200
L = 10_000_000
p = 0.2

r = np.arange(L / p, dtype=int)
s = [
    r[np.random.rand(r.shape[0]) > 1 - p]
    for _ in range(N)

0x0L

#include <algorithm>
#include <iostream>
#include <string>
#include <thread>
#include <vector>
#include <future>
#include <queue>
#include <condition_variable>
#include <mutex>
#include <functional>

0x0L

from pathlib import Path

import numpy as np
import pandas as pd


class RaggedIndexer:
    def __init__(self, counts):
        self.stops = np.r_[0, np.cumsum(counts)]
        self.starts = self.stops[:-1]

0x0L

# https://rockylinux.org/cloud-images/
# AMI 0951577d22b8e8dd4 rocky 9.1
# Instance p3.2xlarge

dnf install -y epel-release
crb enable

dnf config-manager --add-repo https://developer.download.nvidia.com/compute/cuda/repos/rhel9/x86_64/cuda-rhel9.repo
dnf module install -y nvidia-driver:latest
reboot

0x0L

import numpy as np
from opt_einsum import contract

class DRE:
    """Deep Regression Ensemble

    Parameters
    ----------
    n_layers : int
        Number of hidden layers.

0x0L

0x0L

import os

from multiprocessing import get_context
from concurrent.futures import ProcessPoolExecutor
from concurrent.futures.thread import _global_shutdown_lock as L

os.register_at_fork(
    before=L.acquire,
    after_in_child=L._at_fork_reinit,
    after_in_parent=L.release

0x0L

import heapq
import numpy as np

def generic_linkage(D, method='single'):
    N = D.shape[0]
    S = set(range(N))
    
    size = {i: 1 for i in range(N)}

    d = {}

0x0L

import numba
import numpy as np

@numba.jit(nopython=True, fastmath=True)
def move_regress(X, Y, w, fit_intercept=True):
    """Moving window multi-regressions
    
    Solves the least-squares problems `Y[t-w+1:t+1] = X[t-w+1:t+1] @ B[t] + A[t]`
    for `B` and `A` for all `t` in `[w-1, len(Y)-1]`.
 

0x0L