""" getBHv wrapper codes"""
import numpy as np
from scipy.spatial.transform import Rotation as R
from magpylib._lib.fields.field_wrap_BH_level1 import getBH_level1
from magpylib._lib.exceptions import MagpylibBadUserInput
def getBHv_level2(**kwargs: dict) -> np.ndarray:
""" Direct access to vectorized computation
Parameters
----------
kwargs: dict that describes the computation.
Returns
-------
field: ndarray, shape (N,3), field at obs_pos in [mT] or [kA/m]
Info
----
- check inputs
- secures input types (list/tuple -> ndarray)
- test if mandatory inputs are there
- sets default input variables (e.g. pos, rot) if missing
- tiles 1D inputs vectors to correct dimension
"""
# pylint: disable=too-many-branches
# pylint: disable=too-many-statements
# generate dict of secured inputs for auto-tiling ---------------
# entries in this dict will be tested for input length, and then
# be automatically tiled up and stored back into kwargs for calling
# getBH_level1().
# To allow different input dimensions, the tdim argument is also given
# which tells the program which dimension it should tile up.
tile_params = {}
# mandatory general inputs ------------------
try:
src_type = kwargs['source_type']
poso = np.array(kwargs['observer'], dtype=float)
tile_params['observer'] = (poso,2) # <-- (input,tdim)
# optional general inputs -------------------
# if no input set pos=0
pos = np.array(kwargs.get('position', (0,0,0)), dtype=float)
tile_params['position'] = (pos,2)
# if no input set rot=unit
rot = kwargs.get('orientation', R.from_quat((0,0,0,1)))
tile_params['orientation'] = (rot.as_quat(),2)
# if no input set squeeze=True
squeeze = kwargs.get('squeeze', True)
# mandatory class specific inputs -----------
if src_type == 'Box':
mag = np.array(kwargs['magnetization'], dtype=float)
tile_params['magnetization'] = (mag,2)
dim = np.array(kwargs['dimension'], dtype=float)
tile_params['dimension'] = (dim,2)
elif src_type == 'Cylinder':
mag = np.array(kwargs['magnetization'], dtype=float)
tile_params['magnetization'] = (mag,2)
dim = np.array(kwargs['dimension'], dtype=float)
tile_params['dimension'] = (dim,2)
elif src_type == 'Sphere':
mag = np.array(kwargs['magnetization'], dtype=float)
tile_params['magnetization'] = (mag,2)
dia = np.array(kwargs['diameter'], dtype=float)
tile_params['diameter'] = (dia,1)
elif src_type == 'Dipole':
moment = np.array(kwargs['moment'], dtype=float)
tile_params['moment'] = (moment,2)
elif src_type == 'Circular':
current = np.array(kwargs['current'], dtype=float)
tile_params['current'] = (current,1)
dia = np.array(kwargs['diameter'], dtype=float)
tile_params['diameter'] = (dia,1)
elif src_type == 'Line':
current = np.array(kwargs['current'], dtype=float)
tile_params['current'] = (current,1)
pos_start = np.array(kwargs['segment_start'], dtype=float)
tile_params['segment_start'] = (pos_start,2)
pos_end = np.array(kwargs['segment_end'], dtype=float)
tile_params['segment_end'] = (pos_end,2)
except KeyError as kerr:
msg = f'Missing input keys: {str(kerr)}'
raise MagpylibBadUserInput(msg) from kerr
# auto tile 1D parameters ---------------------------------------
# evaluation vector length
ns = [len(val) for val,tdim in tile_params.values() if val.ndim == tdim]
if len(set(ns)) > 1:
msg = f'getBHv() bad array input lengths: {str(set(ns))}'
raise MagpylibBadUserInput(msg)
n = max(ns, default=1)
# tile 1D inputs and replace original values in kwargs
for key,(val,tdim) in tile_params.items():
if val.ndim<tdim:
if tdim == 2:
kwargs[key] = np.tile(val,(n,1))
elif tdim == 1:
kwargs[key] = np.array([val]*n)
else:
kwargs[key] = val
# change rot to Rotation object
kwargs['orientation'] = R.from_quat(kwargs['orientation'])
# compute and return B
B = getBH_level1(**kwargs)
if squeeze:
return np.squeeze(B)
return B
# ON INTERFACE
[docs]def getBv(**kwargs):
"""
B-Field computation in units of [mT] from a dictionary of input vectors of
length N.
This function avoids the object-oriented Magpylib interface and gives direct
access to the field implementations. It is the fastet way to compute fields
with Magpylib.
"Static" inputs of shape (x,) will automatically be tiled to shape (N,x) to
fit with other inputs.
Required inputs depend on chosen source_type!
Parameters
----------
source_type: string
Source type for computation. Must be either 'Box', 'Cylinder', 'Sphere', 'Dipole',
'Circular' or 'Line'. Expected input parameters depend on source_type.
position: array_like, shape (3,) or (N,3), default=(0,0,0)
Source positions in units of [mm].
orientation: scipy Rotation object, default=unit rotation
Source rotations relative to the initial state (see object docstrings).
observer: array_like, shape (3,) or (N,3)
Observer positions in units of [mm].
squeeze: bool, default=True
If True, the output is squeezed, i.e. all axes of length 1 in the output are eliminated.
magnetization: array_like, shape (3,) or (N,3)
Only `source_type in ('Box', 'Cylinder', 'Sphere')`! Magnetization vector (mu0*M) or
remanence field of homogeneous magnet magnetization in units of [mT].
moment: array_like, shape (3,) or (N,3)
Only `source_type = 'Moment'`! Magnetic dipole moment in units of [mT*mm^3]. For
homogeneous magnets the relation is moment = magnetization*volume.
current: array_like, shape (N,)
Only `source_type in ('Line', 'Circular')`! Current flowing in loop in units of [A].
dimension: array_like
Only `source_type in ('Box', 'Cylinder')`! Magnet dimension input in units of [mm].
diameter: array_like, shape (N)
Only `source_type in (Sphere, Circular)`! Diameter of source in units of [mm].
segment_start: array_like, shape (N,3)
Only `source_type = 'Line'`! Start positions of line current segments in units of [mm].
segment_end: array_like, shape (N,3)
Only `source_type = 'Line'`! End positions of line current segments in units of [mm].
Returns
-------
B-field: ndarray, shape (N,3)
B-field generated by sources at observer positions in units of [mT].
Examples
--------
Three-fold evaluation of the dipole field. For each computation the moment is (100,100,100).
>>> import magpylib as mag3
>>> B = mag3.getBv(
>>> source_type='Dipole',
>>> position=[(1,2,3), (2,3,4), (3,4,5)],
>>> moment=(100,100,100),
>>> observer=[(1,1,1), (2,2,2), (3,3,3)])
>>> print(B)
[[-0.71176254 0.56941003 1.85058261]
[-0.71176254 0.56941003 1.85058261]
[-0.71176254 0.56941003 1.85058261]]
Six-fold evaluation of a Cuboid magnet field with increasing size and magnetization
of the magnet. Position and orientation are by default (0,0,0) and unit-orientation,
respectively. The observer position is (1,2,3) for each evaluation.
>>> import numpy as np
>>> import magpylib as mag3
>>> B = mag3.getBv(
>>> source_type='Box',
>>> magnetization = [(0,0,m) for m in np.linspace(500,1000,6)],
>>> dimension = [(a,a,a) for a in np.linspace(1,2,6)],
>>> observer=(1,2,3))
>>> print(B)
[[ 0.48818967 0.97689261 0.70605984]
[ 1.01203491 2.02636222 1.46575704]
[ 1.87397714 3.756164 2.72063422]
[ 3.19414311 6.41330652 4.65485356]
[ 5.10909461 10.2855981 7.4881383 ]
[ 7.76954697 15.70382556 11.48192812]]
"""
return getBHv_level2(bh=True, **kwargs)
# ON INTERFACE
[docs]def getHv(**kwargs):
"""
H-Field computation in units of [kA/m] from a dictionary of input vectors of
length N.
This function avoids the object-oriented Magpylib interface and gives direct
access to the field implementations. It is the fastet way to compute fields
with Magpylib.
"Static" inputs of shape (x,) will automatically be tiled to shape (N,x) to
fit with other inputs.
Required inputs depend on chosen source_type!
Parameters
----------
source_type: string
Source type for computation. Must be either 'Box', 'Cylinder', 'Sphere', 'Dipole',
'Circular' or 'Line'. Expected input parameters depend on source_type.
position: array_like, shape (3,) or (N,3), default=(0,0,0)
Source positions in units of [mm].
orientation: scipy Rotation object, default=unit rotation
Source rotations relative to the initial state (see object docstrings).
observer: array_like, shape (3,) or (N,3)
Observer positions in units of [mm].
squeeze: bool, default=True
If True, the output is squeezed, i.e. all axes of length 1 in the output are eliminated.
magnetization: array_like, shape (3,) or (N,3)
Only `source_type in ('Box', 'Cylinder', 'Sphere')`! Magnetization vector (mu0*M) or
remanence field of homogeneous magnet magnetization in units of [mT].
moment: array_like, shape (3,) or (N,3)
Only `source_type = 'Moment'`! Magnetic dipole moment in units of [mT*mm^3]. For
homogeneous magnets the relation is moment = magnetization*volume.
current: array_like, shape (N,)
Only `source_type in ('Line', 'Circular')`! Current flowing in loop in units of [A].
dimension: array_like
Only `source_type in ('Box', 'Cylinder')`! Magnet dimension input in units of [mm].
diameter: array_like, shape (N)
Only `source_type in (Sphere, Circular)`! Diameter of source in units of [mm].
segment_start: array_like, shape (N,3)
Only `source_type = 'Line'`! Start positions of line current segments in units of [mm].
segment_end: array_like, shape (N,3)
Only `source_type = 'Line'`! End positions of line current segments in units of [mm].
Returns
-------
H-field: ndarray, shape (N,3)
H-field generated by sources at observer positions in units of [kA/m].
Examples
--------
Three-fold evaluation of the dipole field. For each computation the moment is (100,100,100).
>>> import magpylib as mag3
>>> H = mag3.getHv(
>>> source_type='Dipole',
>>> position=[(1,2,3), (2,3,4), (3,4,5)],
>>> moment=(100,100,100),
>>> observer=[(1,1,1), (2,2,2), (3,3,3)])
>>> print(H)
[[-0.56640264 0.45312211 1.47264685]
[-0.56640264 0.45312211 1.47264685]
[-0.56640264 0.45312211 1.47264685]]
Six-fold evaluation of a Cuboid magnet field with increasing size and magnetization
of the magnet. Position and orientation are (0,0,0) and unit-orientation, respectively,
by default. The observer position is (1,2,3) for each evaluation.
>>> import numpy as np
>>> import magpylib as mag3
>>> H = mag3.getHv(
>>> source_type='Box',
>>> magnetization = [(0,0,m) for m in np.linspace(500,1000,6)],
>>> dimension = [(a,a,a) for a in np.linspace(1,2,6)],
>>> observer=(1,2,3))
>>> print(H)
[[ 0.388489 0.77738644 0.56186457]
[ 0.80535179 1.61252782 1.16641239]
[ 1.49126363 2.98906034 2.16501192]
[ 2.54181833 5.10354717 3.70421476]
[ 4.06568831 8.1850189 5.95887113]
[ 6.18280903 12.49670731 9.13702808]]
"""
return getBHv_level2(bh=False, **kwargs)