loganbvh

import h5py
import numpy as np

def set_h5_attrs(grp, data):
    """Sets attributes of h5py group or File `grp` according to dict `data`.
    
    Args:
        grp (h5py group or File): Group or file you would like to update.
        data (dict): Dict of data with which to update `grp`.
    """

loganbvh

from numpy import sqrt
from scipy.special import ellipe, ellipk
import pint

ureg = pint.UnitRegistry()

def mutual_inductance(r1: float, r2: float, h: float, length_units: str = "m") -> pint.Quantity:
    """Calculates the mutual inductance between two coaxial circular loops.
    
    All lengths, ``r1``, ``r2``, and ``h`` are assumed to be given in ``length_units``.

loganbvh

import os

import requests  # pip install requests
from tqdm import tqdm  # pip install tqdm
from tqdm.utils import CallbackIOWrapper

# See: https://gist.github.com/slint/92e4d38eb49dd177f46b02e1fe9761e1
# See: https://gist.github.com/tyhoff/b757e6af83c1fd2b7b83057adf02c139

def get_bucket_url(deposit_id: str, access_token: str) -> str:

loganbvh

import numba
import numpy as np

@numba.njit(fastmath=True, parallel=True)
def fast_matmul(A: np.ndarray, B: np.ndarray) -> np.ndarray:
    """Performs ``A @ B`` for 2D matrices efficiently using numba."""
    # I have found that pre-allocating ``out`` and passing it in as an argument does
    # not speed things up very much.
    # On my M1 MacBook with 10 cores, this function is faster and has lower CPU
    # utilization than A @ B using numpy/BLAS. It is also slightly faster than

loganbvh

loganbvh

import numpy as np
import superscreen as sc

def field_from_solution(
    x: np.ndarray,
    y: np.ndarray,
    z: np.ndarray,
    *,
    solution: sc.Solution,
    dr: tuple[float, float, float] = (0.0, 0.0, 0.0),

loganbvh

import os
import sys

import numpy as np
import pint
from scipy.spatial.transform import Rotation
import superscreen as sc
from superscreen.geometry import box

sys.path.insert(0, os.path.expanduser("~/GitHub/superscreen-squids"))

loganbvh