DAMASK_EICMD/python/damask/util.py

756 lines
22 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

"""Miscellaneous helper functionality."""
import sys
import datetime
import os
import subprocess
import shlex
import re
import fractions
import collections.abc as abc
from functools import reduce
from typing import Union, Tuple, Iterable, Sequence, Callable, Dict, List, Any, Literal, Optional
from pathlib import Path
import numpy as np
import h5py
from . import version
from ._typehints import FloatSequence, IntSequence
# limit visibility
__all__=[
'srepr',
'emph', 'deemph', 'warn', 'strikeout',
'run',
'natural_sort',
'show_progress',
'scale_to_coprime',
'project_equal_angle', 'project_equal_area',
'hybrid_IA',
'execution_stamp',
'shapeshifter', 'shapeblender',
'extend_docstring', 'extended_docstring',
'Bravais_to_Miller', 'Miller_to_Bravais',
'DREAM3D_base_group', 'DREAM3D_cell_data_group',
'dict_prune', 'dict_flatten'
]
# https://svn.blender.org/svnroot/bf-blender/trunk/blender/build_files/scons/tools/bcolors.py
# https://stackoverflow.com/questions/287871
_colors = {
'header' : '\033[95m',
'OK_blue': '\033[94m',
'OK_green': '\033[92m',
'warning': '\033[93m',
'fail': '\033[91m',
'end_color': '\033[0m',
'bold': '\033[1m',
'dim': '\033[2m',
'underline': '\033[4m',
'crossout': '\033[9m'
}
####################################################################################################
# Functions
####################################################################################################
def srepr(msg, glue: str = '\n') -> str:
r"""
Join items with glue string.
Parameters
----------
msg : object with __repr__ or sequence of objects with __repr__
Items to join.
glue : str, optional
Glue used for joining operation. Defaults to '\n'.
Returns
-------
joined : str
String representation of the joined items.
"""
if (not hasattr(msg, 'strip') and
(hasattr(msg, '__getitem__') or
hasattr(msg, '__iter__'))):
return glue.join(str(x) for x in msg)
else:
return msg if isinstance(msg,str) else repr(msg)
def emph(msg) -> str:
"""
Format with emphasis.
Parameters
----------
msg : object with __repr__ or sequence of objects with __repr__
Message to format.
Returns
-------
formatted : str
Formatted string representation of the joined items.
"""
return _colors['bold']+srepr(msg)+_colors['end_color']
def deemph(msg) -> str:
"""
Format with deemphasis.
Parameters
----------
msg : object with __repr__ or sequence of objects with __repr__
Message to format.
Returns
-------
formatted : str
Formatted string representation of the joined items.
"""
return _colors['dim']+srepr(msg)+_colors['end_color']
def warn(msg) -> str:
"""
Format for warning.
Parameters
----------
msg : object with __repr__ or sequence of objects with __repr__
Message to format.
Returns
-------
formatted : str
Formatted string representation of the joined items.
"""
return _colors['warning']+emph(msg)+_colors['end_color']
def strikeout(msg) -> str:
"""
Format as strikeout.
Parameters
----------
msg : object with __repr__ or iterable of objects with __repr__
Message to format.
Returns
-------
formatted : str
Formatted string representation of the joined items.
"""
return _colors['crossout']+srepr(msg)+_colors['end_color']
def run(cmd: str, wd: str = './', env: Dict[str, str] = None, timeout: int = None) -> Tuple[str, str]:
"""
Run a command.
Parameters
----------
cmd : str
Command to be executed.
wd : str, optional
Working directory of process. Defaults to './'.
env : dict, optional
Environment for execution.
timeout : integer, optional
Timeout in seconds.
Returns
-------
stdout, stderr : (str, str)
Output of the executed command.
"""
print(f"running '{cmd}' in '{wd}'")
process = subprocess.run(shlex.split(cmd),
stdout = subprocess.PIPE,
stderr = subprocess.PIPE,
env = os.environ if env is None else env,
cwd = wd,
encoding = 'utf-8',
timeout = timeout)
if process.returncode != 0:
print(process.stdout)
print(process.stderr)
raise RuntimeError(f"'{cmd}' failed with returncode {process.returncode}")
return process.stdout, process.stderr
execute = run
def natural_sort(key: str) -> List[Union[int, str]]:
"""
Natural sort.
For use in python's 'sorted'.
References
----------
https://en.wikipedia.org/wiki/Natural_sort_order
"""
convert = lambda text: int(text) if text.isdigit() else text
return [ convert(c) for c in re.split('([0-9]+)', key) ]
def show_progress(iterable: Iterable,
N_iter: int = None,
prefix: str = '',
bar_length: int = 50) -> Any:
"""
Decorate a loop with a progress bar.
Use similar like enumerate.
Parameters
----------
iterable : iterable
Iterable to be decorated.
N_iter : int, optional
Total number of iterations. Required if iterable is not a sequence.
prefix : str, optional
Prefix string.
bar_length : int, optional
Length of progress bar in characters. Defaults to 50.
"""
if isinstance(iterable,abc.Sequence):
if N_iter is None:
N = len(iterable)
else:
raise ValueError('N_iter given for sequence')
else:
if N_iter is None:
raise ValueError('N_iter not given')
else:
N = N_iter
if N <= 1:
for item in iterable:
yield item
else:
status = ProgressBar(N,prefix,bar_length)
for i,item in enumerate(iterable):
yield item
status.update(i)
def scale_to_coprime(v: FloatSequence) -> np.ndarray:
"""
Scale vector to co-prime (relatively prime) integers.
Parameters
----------
v : sequence of float, len (:)
Vector to scale.
Returns
-------
m : numpy.ndarray, shape (:)
Vector scaled to co-prime numbers.
"""
MAX_DENOMINATOR = 1000000
def get_square_denominator(x):
"""Denominator of the square of a number."""
return fractions.Fraction(x ** 2).limit_denominator(MAX_DENOMINATOR).denominator
def lcm(a,b):
"""Least common multiple."""
try:
return np.lcm(a,b) # numpy > 1.18
except AttributeError:
return a * b // np.gcd(a, b)
v_ = np.array(v)
m = (v_ * reduce(lcm, map(lambda x: int(get_square_denominator(x)),v_))**0.5).astype(int)
m = m//reduce(np.gcd,m)
with np.errstate(invalid='ignore'):
if not np.allclose(np.ma.masked_invalid(v_/m),v_[np.argmax(abs(v_))]/m[np.argmax(abs(v_))]):
raise ValueError(f'Invalid result {m} for input {v_}. Insufficient precision?')
return m
def project_equal_angle(vector: np.ndarray,
direction: Literal['x', 'y', 'z'] = 'z',
normalize: bool = True,
keepdims: bool = False) -> np.ndarray:
"""
Apply equal-angle projection to vector.
Parameters
----------
vector : numpy.ndarray, shape (...,3)
Vector coordinates to be projected.
direction : {'x', 'y', 'z'}
Projection direction. Defaults to 'z'.
normalize : bool
Ensure unit length of input vector. Defaults to True.
keepdims : bool
Maintain three-dimensional output coordinates. Defaults to False.
Two-dimensional output uses right-handed frame spanned by
the next and next-next axis relative to the projection direction,
e.g. x-y when projecting along z and z-x when projecting along y.
Returns
-------
coordinates : numpy.ndarray, shape (...,2 | 3)
Projected coordinates.
Examples
--------
>>> import damask
>>> import numpy as np
>>> project_equal_angle(np.ones(3))
[0.3660254, 0.3660254]
>>> project_equal_angle(np.ones(3),direction='x',normalize=False,keepdims=True)
[0, 0.5, 0.5]
>>> project_equal_angle([0,1,1],direction='y',normalize=True,keepdims=False)
[0.41421356, 0]
"""
shift = 'zyx'.index(direction)
v = np.roll(vector/np.linalg.norm(vector,axis=-1,keepdims=True) if normalize else vector,
shift,axis=-1)
return np.roll(np.block([v[...,:2]/(1.0+np.abs(v[...,2:3])),np.zeros_like(v[...,2:3])]),
-shift if keepdims else 0,axis=-1)[...,:3 if keepdims else 2]
def project_equal_area(vector: np.ndarray,
direction: Literal['x', 'y', 'z'] = 'z',
normalize: bool = True,
keepdims: bool = False) -> np.ndarray:
"""
Apply equal-area projection to vector.
Parameters
----------
vector : numpy.ndarray, shape (...,3)
Vector coordinates to be projected.
direction : {'x', 'y', 'z'}
Projection direction. Defaults to 'z'.
normalize : bool
Ensure unit length of input vector. Defaults to True.
keepdims : bool
Maintain three-dimensional output coordinates. Defaults to False.
Two-dimensional output uses right-handed frame spanned by
the next and next-next axis relative to the projection direction,
e.g. x-y when projecting along z and z-x when projecting along y.
Returns
-------
coordinates : numpy.ndarray, shape (...,2 | 3)
Projected coordinates.
Examples
--------
>>> import damask
>>> import numpy as np
>>> project_equal_area(np.ones(3))
[0.45970084, 0.45970084]
>>> project_equal_area(np.ones(3),direction='x',normalize=False,keepdims=True)
[0.0, 0.70710678, 0.70710678]
>>> project_equal_area([0,1,1],direction='y',normalize=True,keepdims=False)
[0.5411961, 0.0]
"""
shift = 'zyx'.index(direction)
v = np.roll(vector/np.linalg.norm(vector,axis=-1,keepdims=True) if normalize else vector,
shift,axis=-1)
return np.roll(np.block([v[...,:2]/np.sqrt(1.0+np.abs(v[...,2:3])),np.zeros_like(v[...,2:3])]),
-shift if keepdims else 0,axis=-1)[...,:3 if keepdims else 2]
def execution_stamp(class_name: str, function_name: str = None) -> str:
"""Timestamp the execution of a (function within a) class."""
now = datetime.datetime.now().astimezone().strftime('%Y-%m-%d %H:%M:%S%z')
_function_name = '' if function_name is None else f'.{function_name}'
return f'damask.{class_name}{_function_name} v{version} ({now})'
def hybrid_IA(dist: np.ndarray, N: int, rng_seed = None) -> np.ndarray:
"""
Hybrid integer approximation.
Parameters
----------
dist : numpy.ndarray
Distribution to be approximated
N : int
Number of samples to draw.
rng_seed : {None, int, array_like[ints], SeedSequence, BitGenerator, Generator}, optional
A seed to initialize the BitGenerator. Defaults to None.
If None, then fresh, unpredictable entropy will be pulled from the OS.
"""
N_opt_samples,N_inv_samples = (max(np.count_nonzero(dist),N),0) # random subsampling if too little samples requested
scale_,scale,inc_factor = (0.0,float(N_opt_samples),1.0)
while (not np.isclose(scale, scale_)) and (N_inv_samples != N_opt_samples):
repeats = np.rint(scale*dist).astype(int)
N_inv_samples = np.sum(repeats)
scale_,scale,inc_factor = (scale,scale+inc_factor*0.5*(scale - scale_), inc_factor*2.0) \
if N_inv_samples < N_opt_samples else \
(scale_,0.5*(scale_ + scale), 1.0)
return np.repeat(np.arange(len(dist)),repeats)[np.random.default_rng(rng_seed).permutation(N_inv_samples)[:N]]
def shapeshifter(fro: Tuple[int, ...],
to: Tuple[int, ...],
mode: Literal['left','right'] = 'left',
keep_ones: bool = False) -> Tuple[Optional[int], ...]:
"""
Return dimensions that reshape 'fro' to become broadcastable to 'to'.
Parameters
----------
fro : tuple
Original shape of array.
to : tuple
Target shape of array after broadcasting.
len(to) cannot be less than len(fro).
mode : {'left', 'right'}, optional
Indicates whether new axes are preferably added to
either left or right of the original shape.
Defaults to 'left'.
keep_ones : bool, optional
Treat '1' in fro as literal value instead of dimensional placeholder.
Defaults to False.
Returns
-------
new_dims : tuple
Dimensions for reshape.
Example
-------
>>> import numpy as np
>>> from damask import util
>>> a = np.ones((3,4,2))
>>> b = np.ones(4)
>>> b_extended = b.reshape(util.shapeshifter(b.shape,a.shape))
>>> (a * np.broadcast_to(b_extended,a.shape)).shape
(3,4,2)
"""
if not len(fro) and not len(to): return ()
beg = dict(left ='(^.*\\b)',
right='(^.*?\\b)')
sep = dict(left ='(.*\\b)',
right='(.*?\\b)')
end = dict(left ='(.*?$)',
right='(.*$)')
fro = (1,) if not len(fro) else fro
to = (1,) if not len(to) else to
try:
match = re.match(beg[mode]
+f',{sep[mode]}'.join(map(lambda x: f'{x}'
if x>1 or (keep_ones and len(fro)>1) else
'\\d+',fro))
+f',{end[mode]}',','.join(map(str,to))+',')
assert match
grp = match.groups()
except AssertionError:
raise ValueError(f'Shapes can not be shifted {fro} --> {to}')
fill: Tuple[Optional[int], ...] = ()
for g,d in zip(grp,fro+(None,)):
fill += (1,)*g.count(',')+(d,)
return fill[:-1]
def shapeblender(a: Tuple[int, ...], b: Tuple[int, ...]) -> Tuple[int, ...]:
"""
Return a shape that overlaps the rightmost entries of 'a' with the leftmost of 'b'.
Parameters
----------
a : tuple
Shape of first array.
b : tuple
Shape of second array.
Examples
--------
>>> shapeblender((4,4,3),(3,2,1))
(4,4,3,2,1)
>>> shapeblender((1,2),(1,2,3))
(1,2,3)
>>> shapeblender((1,),(2,2,1))
(1,2,2,1)
>>> shapeblender((3,2),(3,2))
(3,2)
"""
i = min(len(a),len(b))
while i > 0 and a[-i:] != b[:i]: i -= 1
return a + b[i:]
def extend_docstring(extra_docstring: str) -> Callable:
"""
Decorator: Append to function's docstring.
Parameters
----------
extra_docstring : str
Docstring to append.
"""
def _decorator(func):
func.__doc__ += extra_docstring
return func
return _decorator
def extended_docstring(f: Callable, extra_docstring: str) -> Callable:
"""
Decorator: Combine another function's docstring with a given docstring.
Parameters
----------
f : function
Function of which the docstring is taken.
extra_docstring : str
Docstring to append.
"""
def _decorator(func):
func.__doc__ = f.__doc__ + extra_docstring
return func
return _decorator
def DREAM3D_base_group(fname: Union[str, Path]) -> str:
"""
Determine the base group of a DREAM.3D file.
The base group is defined as the group (folder) that contains
a 'SPACING' dataset in a '_SIMPL_GEOMETRY' group.
Parameters
----------
fname : str or pathlib.Path
Filename of the DREAM.3D (HDF5) file.
Returns
-------
path : str
Path to the base group.
"""
with h5py.File(fname,'r') as f:
base_group = f.visit(lambda path: path.rsplit('/',2)[0] if '_SIMPL_GEOMETRY/SPACING' in path else None)
if base_group is None:
raise ValueError(f'Could not determine base group in file {fname}.')
return base_group
def DREAM3D_cell_data_group(fname: Union[str, Path]) -> str:
"""
Determine the cell data group of a DREAM.3D file.
The cell data group is defined as the group (folder) that contains
a dataset in the base group whose length matches the total number
of points as specified in '_SIMPL_GEOMETRY/DIMENSIONS'.
Parameters
----------
fname : str or pathlib.Path
Filename of the DREAM.3D (HDF5) file.
Returns
-------
path : str
Path to the cell data group.
"""
base_group = DREAM3D_base_group(fname)
with h5py.File(fname,'r') as f:
cells = tuple(f['/'.join([base_group,'_SIMPL_GEOMETRY','DIMENSIONS'])][()][::-1])
cell_data_group = f[base_group].visititems(lambda path,obj: path.split('/')[0] \
if isinstance(obj,h5py._hl.dataset.Dataset) and np.shape(obj)[:-1] == cells \
else None)
if cell_data_group is None:
raise ValueError(f'Could not determine cell data group in file {fname}/{base_group}.')
return cell_data_group
def Bravais_to_Miller(*, uvtw: np.ndarray = None, hkil: np.ndarray = None) -> np.ndarray:
"""
Transform 4 MillerBravais indices to 3 Miller indices of crystal direction [uvw] or plane normal (hkl).
Parameters
----------
uvtw|hkil : numpy.ndarray, shape (...,4)
MillerBravais indices of crystallographic direction [uvtw] or plane normal (hkil).
Returns
-------
uvw|hkl : numpy.ndarray, shape (...,3)
Miller indices of [uvw] direction or (hkl) plane normal.
"""
if (uvtw is not None) ^ (hkil is None):
raise KeyError('Specify either "uvtw" or "hkil"')
axis,basis = (np.array(uvtw),np.array([[1,0,-1,0],
[0,1,-1,0],
[0,0, 0,1]])) \
if hkil is None else \
(np.array(hkil),np.array([[1,0,0,0],
[0,1,0,0],
[0,0,0,1]]))
return np.einsum('il,...l',basis,axis)
def Miller_to_Bravais(*, uvw: np.ndarray = None, hkl: np.ndarray = None) -> np.ndarray:
"""
Transform 3 Miller indices to 4 MillerBravais indices of crystal direction [uvtw] or plane normal (hkil).
Parameters
----------
uvw|hkl : numpy.ndarray, shape (...,3)
Miller indices of crystallographic direction [uvw] or plane normal (hkl).
Returns
-------
uvtw|hkil : numpy.ndarray, shape (...,4)
MillerBravais indices of [uvtw] direction or (hkil) plane normal.
"""
if (uvw is not None) ^ (hkl is None):
raise KeyError('Specify either "uvw" or "hkl"')
axis,basis = (np.array(uvw),np.array([[ 2,-1, 0],
[-1, 2, 0],
[-1,-1, 0],
[ 0, 0, 3]])/3) \
if hkl is None else \
(np.array(hkl),np.array([[ 1, 0, 0],
[ 0, 1, 0],
[-1,-1, 0],
[ 0, 0, 1]]))
return np.einsum('il,...l',basis,axis)
def dict_prune(d: Dict) -> Dict:
"""
Recursively remove empty dictionaries.
Parameters
----------
d : dict
Dictionary to prune.
Returns
-------
pruned : dict
Pruned dictionary.
"""
# https://stackoverflow.com/questions/48151953
new = {}
for k,v in d.items():
if isinstance(v, dict):
v = dict_prune(v)
if not isinstance(v,dict) or v != {}:
new[k] = v
return new
def dict_flatten(d: Dict) -> Dict:
"""
Recursively remove keys of single-entry dictionaries.
Parameters
----------
d : dict
Dictionary to flatten.
Returns
-------
flattened : dict
Flattened dictionary.
"""
if isinstance(d,dict) and len(d) == 1:
entry = d[list(d.keys())[0]]
new = dict_flatten(entry.copy()) if isinstance(entry,dict) else entry
else:
new = {k: (dict_flatten(v) if isinstance(v, dict) else v) for k,v in d.items()}
return new
####################################################################################################
# Classes
####################################################################################################
class ProgressBar:
"""
Report progress of an interation as a status bar.
Works for 0-based loops, ETA is estimated by linear extrapolation.
"""
def __init__(self, total: int, prefix: str, bar_length: int):
"""
Set current time as basis for ETA estimation.
Parameters
----------
total : int
Total # of iterations.
prefix : str
Prefix string.
bar_length : int
Character length of bar.
"""
self.total = total
self.prefix = prefix
self.bar_length = bar_length
self.time_start = self.time_last_update = datetime.datetime.now()
self.fraction_last = 0.0
sys.stderr.write(f"{self.prefix} {''*self.bar_length} 0% ETA n/a")
sys.stderr.flush()
def update(self, iteration: int) -> None:
fraction = (iteration+1) / self.total
filled_length = int(self.bar_length * fraction)
if filled_length > int(self.bar_length * self.fraction_last) or \
datetime.datetime.now() - self.time_last_update > datetime.timedelta(seconds=10):
self.time_last_update = datetime.datetime.now()
bar = '' * filled_length + '' * (self.bar_length - filled_length)
remaining_time = (datetime.datetime.now() - self.time_start) \
* (self.total - (iteration+1)) / (iteration+1)
remaining_time -= datetime.timedelta(microseconds=remaining_time.microseconds) # remove μs
sys.stderr.write(f'\r{self.prefix} {bar} {fraction:>4.0%} ETA {remaining_time}')
sys.stderr.flush()
self.fraction_last = fraction
if iteration == self.total - 1:
sys.stderr.write('\n')
sys.stderr.flush()