def issubdtype(arg1, arg2):
"""
Returns True if first argument is a typecode lower/equal in type hierarchy.
Parameters
----------
arg1, arg2 : dtype_like
dtype or string representing a typecode.
Returns
-------
out : bool
See Also
--------
issubsctype, issubclass_
numpy.core.numerictypes : Overview of numpy type hierarchy.
Examples
--------
>>> np.issubdtype('S1', np.string_)
True
>>> np.issubdtype(np.float64, np.float32)
False
"""
if not issubclass_(arg1, generic):
arg1 = dtype(arg1).type
if not issubclass_(arg2, generic):
arg2_orig = arg2
arg2 = dtype(arg2).type
if not isinstance(arg2_orig, dtype):
# weird deprecated behaviour, that tried to infer np.floating from
# float, and similar less obvious things, such as np.generic from
# basestring
mro = arg2.mro()
arg2 = mro[1] if len(mro) > 1 else mro[0]
def type_repr(x):
""" Helper to produce clear error messages """
if not isinstance(x, type):
return repr(x)
elif issubclass(x, generic):
return "np.{}".format(x.__name__)
else:
return x.__name__
# 1.14, 2017-08-01
warnings.warn(
"Conversion of the second argument of issubdtype from `{raw}` "
"to `{abstract}` is deprecated. In future, it will be treated "
"as `{concrete} == np.dtype({raw}).type`.".format(
raw=type_repr(arg2_orig),
abstract=type_repr(arg2),
concrete=type_repr(dtype(arg2_orig).type)
),
FutureWarning, stacklevel=2
)
return issubclass(arg1, arg2)
def bincount(x, weights=None, minlength=None):
if minlength is None:
minlength = 0
else:
if not isinstance(minlength, (int, long)):
raise TypeError("an integer is required")
if minlength <= 0:
raise ValueError("minlength must be positive")
x = array(x)
len_output = minlength
if len(x) > 0:
if x.min() < 0:
raise ValueError("x must not be negative")
len_output = max(len_output, x.max() + 1)
if x.dtype.kind not in 'ui':
raise ValueError("x must be integer")
if weights is None:
output = zeros(len_output, dtype=dtype('int'))
for elem in x:
output[elem] += 1
else:
if len(weights) != len(x):
raise ValueError("x and weights arrays must have the same size")
output = zeros(len_output, dtype=dtype('float'))
for i in range(len(x)):
output[x[i]] += weights[i]
return output
def _add_trailing_padding(value, padding):
"""Inject the specified number of padding bytes at the end of a dtype"""
from numpy.core.multiarray import dtype
if value.fields is None:
vfields = {'f0': (value, 0)}
else:
vfields = dict(value.fields)
if value.names and value.names[-1] == '' and \
value[''].char == 'V':
# A trailing padding field is already present
vfields[''] = ('V%d' % (vfields[''][0].itemsize + padding),
vfields[''][1])
value = dtype(vfields)
else:
# Get a free name for the padding field
j = 0
while True:
name = 'pad%d' % j
if name not in vfields:
vfields[name] = ('V%d' % padding, value.itemsize)
break
j += 1
value = dtype(vfields)
if '' not in vfields:
# Strip out the name of the padding field
names = list(value.names)
names[-1] = ''
value.names = tuple(names)
return value
def _mean(a, axis=None, dtype=None, out=None, keepdims=False):
arr = asanyarray(a)
is_float16_result = False
rcount = _count_reduce_items(arr, axis)
# Make this warning show up first
if rcount == 0:
warnings.warn("Mean of empty slice.", RuntimeWarning, stacklevel=2)
# Cast bool, unsigned int, and int to float64 by default
if dtype is None:
if issubclass(arr.dtype.type, (nt.integer, nt.bool_)):
dtype = mu.dtype('f8')
elif issubclass(arr.dtype.type, nt.float16):
dtype = mu.dtype('f4')
is_float16_result = True
ret = umr_sum(arr, axis, dtype, out, keepdims)
if isinstance(ret, mu.ndarray):
ret = um.true_divide(
ret, rcount, out=ret, casting='unsafe', subok=False)
if is_float16_result and out is None:
ret = arr.dtype.type(ret)
elif hasattr(ret, 'dtype'):
if is_float16_result:
ret = arr.dtype.type(ret / rcount)
else:
ret = ret.dtype.type(ret / rcount)
else:
ret = ret / rcount
return ret
def issubdtype(arg1, arg2):
"""
Returns True if first argument is a typecode lower/equal in type hierarchy.
Parameters
----------
arg1, arg2 : dtype_like
dtype or string representing a typecode.
Returns
-------
out : bool
See Also
--------
issubsctype, issubclass_
numpy.core.numerictypes : Overview of numpy type hierarchy.
Examples
--------
>>> np.issubdtype('S1', str)
True
>>> np.issubdtype(np.float64, np.float32)
False
"""
if issubclass_(arg2, generic):
return issubclass(dtype(arg1).type, arg2)
mro = dtype(arg2).type.mro()
if len(mro) > 1:
val = mro[1]
else:
val = mro[0]
return issubclass(dtype(arg1).type, val)
def find_common_type(array_types, scalar_types):
"""
Determine common type following standard coercion rules
Parameters
----------
array_types : sequence
A list of dtype convertible objects representing arrays
scalar_types : sequence
A list of dtype convertible objects representing scalars
Returns
-------
datatype : dtype
The common data-type which is the maximum of the array_types
ignoring the scalar_types unless the maximum of the scalar_types
is of a different kind.
If the kinds is not understood, then None is returned.
See Also
--------
dtype
"""
array_types = [dtype(x) for x in array_types]
scalar_types = [dtype(x) for x in scalar_types]
if len(scalar_types) == 0:
if len(array_types) == 0:
return None
else:
return max(array_types)
if len(array_types) == 0:
return max(scalar_types)
maxa = max(array_types)
maxsc = max(scalar_types)
try:
index_a = _kind_list.index(maxa.kind)
index_sc = _kind_list.index(maxsc.kind)
except ValueError:
return None
if index_sc > index_a:
return _find_common_coerce(maxsc,maxa)
else:
return maxa
def _makenames_list(adict, align):
from .multiarray import dtype
allfields = []
fnames = list(adict.keys())
for fname in fnames:
obj = adict[fname]
n = len(obj)
if not isinstance(obj, tuple) or n not in [2,3]:
raise ValueError("entry not a 2- or 3- tuple")
if (n > 2) and (obj[2] == fname):
continue
num = int(obj[1])
if (num < 0):
raise ValueError("invalid offset.")
format = dtype(obj[0], align=align)
if (format.itemsize == 0):
raise ValueError("all itemsizes must be fixed.")
if (n > 2):
title = obj[2]
else:
title = None
allfields.append((fname, format, num, title))
# sort by offsets
allfields.sort(key=lambda x: x[2])
names = [x[0] for x in allfields]
formats = [x[1] for x in allfields]
offsets = [x[2] for x in allfields]
titles = [x[3] for x in allfields]
return names, formats, offsets, titles
def _usefields(adict, align):
from .multiarray import dtype
try:
names = adict[-1]
except KeyError:
names = None
if names is None:
names, formats, offsets, titles = _makenames_list(adict, align)
else:
formats = []
offsets = []
titles = []
for name in names:
res = adict[name]
formats.append(res[0])
offsets.append(res[1])
if (len(res) > 2):
titles.append(res[2])
else:
titles.append(None)
return dtype({"names" : names,
"formats" : formats,
"offsets" : offsets,
"titles" : titles}, align)
def obj2sctype(rep, default=None):
"""
Return the scalar dtype or NumPy equivalent of Python type of an object.
Parameters
----------
rep : any
The object of which the type is returned.
default : any, optional
If given, this is returned for objects whose types can not be
determined. If not given, None is returned for those objects.
Returns
-------
dtype : dtype or Python type
The data type of `rep`.
See Also
--------
sctype2char, issctype, issubsctype, issubdtype, maximum_sctype
Examples
--------
>>> np.obj2sctype(np.int32)
<type 'numpy.int32'>
>>> np.obj2sctype(np.array([1., 2.]))
<type 'numpy.float64'>
>>> np.obj2sctype(np.array([1.j]))
<type 'numpy.complex128'>
>>> np.obj2sctype(dict)
<type 'numpy.object_'>
>>> np.obj2sctype('string')
<type 'numpy.string_'>
>>> np.obj2sctype(1, default=list)
<type 'list'>
"""
try:
if issubclass(rep, generic):
return rep
except TypeError:
pass
if isinstance(rep, dtype):
return rep.type
if isinstance(rep, type):
return _python_type(rep)
if isinstance(rep, ndarray):
return rep.dtype.type
try:
res = dtype(rep)
except:
return default
return res.type
def _set_array_types():
ibytes = [1, 2, 4, 8, 16, 32, 64]
fbytes = [2, 4, 8, 10, 12, 16, 32, 64]
for bytes in ibytes:
bits = 8 * bytes
_add_array_type("int", bits)
_add_array_type("uint", bits)
for bytes in fbytes:
bits = 8 * bytes
_add_array_type("float", bits)
_add_array_type("complex", 2 * bits)
_gi = dtype("p")
if _gi.type not in sctypes["int"]:
indx = 0
sz = _gi.itemsize
_lst = sctypes["int"]
while indx < len(_lst) and sz >= _lst[indx](0).itemsize:
indx += 1
sctypes["int"].insert(indx, _gi.type)
sctypes["uint"].insert(indx, dtype("P").type)
def _check_field_overlap(new_fields, old_fields):
""" Perform object memory overlap tests for two data-types (see
_view_is_safe).
This function checks that new fields only access memory contained in old
fields, and that non-object fields are not interpreted as objects and vice
versa.
Parameters
----------
new_fields : list of (data-type, int) pairs
Flat list of (dtype, byte offset) pairs for the new data type, as
returned by _get_all_field_offsets.
old_fields: list of (data-type, int) pairs
Flat list of (dtype, byte offset) pairs for the old data type, as
returned by _get_all_field_offsets.
Raises
------
TypeError
If the new fields are incompatible with the old fields
"""
from .numerictypes import object_
from .multiarray import dtype
#first go byte by byte and check we do not access bytes not in old_fields
new_bytes = set()
for tp, off in new_fields:
new_bytes.update(set(range(off, off+tp.itemsize)))
old_bytes = set()
for tp, off in old_fields:
old_bytes.update(set(range(off, off+tp.itemsize)))
if new_bytes.difference(old_bytes):
raise TypeError("view would access data parent array doesn't own")
#next check that we do not interpret non-Objects as Objects, and vv
obj_offsets = [off for (tp, off) in old_fields if tp.type is object_]
obj_size = dtype(object_).itemsize
for fld_dtype, fld_offset in new_fields:
if fld_dtype.type is object_:
# check we do not create object views where
# there are no objects.
if fld_offset not in obj_offsets:
raise TypeError("cannot view non-Object data as Object type")
else:
# next check we do not create non-object views
# where there are already objects.
# see validate_object_field_overlap for a similar computation.
for obj_offset in obj_offsets:
if (fld_offset < obj_offset + obj_size and
obj_offset < fld_offset + fld_dtype.itemsize):
raise TypeError("cannot view Object as non-Object type")
def _bits_of(obj):
try:
info = next(v for v in _concrete_typeinfo.values() if v.type is obj)
except StopIteration:
if obj in _abstract_types.values():
raise ValueError("Cannot count the bits of an abstract type")
# some third-party type - make a best-guess
return dtype(obj).itemsize * 8
else:
return info.bits
def _set_array_types():
ibytes = [1, 2, 4, 8, 16, 32, 64]
fbytes = [2, 4, 8, 10, 12, 16, 32, 64]
for bytes in ibytes:
bits = 8*bytes
_add_array_type('int', bits)
_add_array_type('uint', bits)
for bytes in fbytes:
bits = 8*bytes
_add_array_type('float', bits)
_add_array_type('complex', 2*bits)
_gi = dtype('p')
if _gi.type not in sctypes['int']:
indx = 0
sz = _gi.itemsize
_lst = sctypes['int']
while (indx < len(_lst) and sz >= _lst[indx](0).itemsize):
indx += 1
sctypes['int'].insert(indx, _gi.type)
sctypes['uint'].insert(indx, dtype('P').type)
def obj2sctype(rep, default=None):
"""
Return the scalar dtype or NumPy equivalent of Python type of an object.
Parameters
----------
rep : any
The object of which the type is returned.
default : any, optional
If given, this is returned for objects whose types can not be
determined. If not given, None is returned for those objects.
Returns
-------
dtype : dtype or Python type
The data type of `rep`.
See Also
--------
sctype2char, issctype, issubsctype, issubdtype, maximum_sctype
Examples
--------
>>> np.obj2sctype(np.int32)
<type 'numpy.int32'>
>>> np.obj2sctype(np.array([1., 2.]))
<type 'numpy.float64'>
>>> np.obj2sctype(np.array([1.j]))
<type 'numpy.complex128'>
>>> np.obj2sctype(dict)
<type 'numpy.object_'>
>>> np.obj2sctype('string')
<type 'numpy.string_'>
>>> np.obj2sctype(1, default=list)
<type 'list'>
"""
# prevent abtract classes being upcast
if isinstance(rep, type) and issubclass(rep, generic):
return rep
# extract dtype from arrays
if isinstance(rep, ndarray):
return rep.dtype.type
# fall back on dtype to convert
try:
res = dtype(rep)
except Exception:
return default
else:
return res.type
def obj2sctype(rep, default=None):
"""
Return the scalar dtype or NumPy equivalent of Python type of an object.
Parameters
----------
rep : any
The object of which the type is returned.
default : any, optional
If given, this is returned for objects whose types can not be
determined. If not given, None is returned for those objects.
Returns
-------
dtype : dtype or Python type
The data type of `rep`.
See Also
--------
sctype2char, issctype, issubsctype, issubdtype, maximum_sctype
Examples
--------
>>> np.obj2sctype(np.int32)
<type 'numpy.int32'>
>>> np.obj2sctype(np.array([1., 2.]))
<type 'numpy.float64'>
>>> np.obj2sctype(np.array([1.j]))
<type 'numpy.complex128'>
>>> np.obj2sctype(dict)
<type 'numpy.object_'>
>>> np.obj2sctype('string')
<type 'numpy.string_'>
>>> np.obj2sctype(1, default=list)
<type 'list'>
"""
# prevent abtract classes being upcast
if isinstance(rep, type) and issubclass(rep, generic):
return rep
# extract dtype from arrays
if isinstance(rep, ndarray):
return rep.dtype.type
# fall back on dtype to convert
try:
res = dtype(rep)
except Exception:
return default
else:
return res.type
def _find_common_coerce(a, b):
if a > b:
return a
try:
thisind = __test_types.index(a.char)
except ValueError:
return None
while thisind < __len_test_types:
newdtype = dtype(__test_types[thisind])
if newdtype >= b and newdtype >= a:
return newdtype
thisind += 1
return None
def bitname(obj):
"""Return a bit-width name for a given type object"""
bits = _bits_of(obj)
char = dtype(obj).kind
base = _kind_to_stem[char]
if base == 'object':
bits = 0
if bits != 0:
char = "%s%d" % (char, bits // 8)
return base, bits, char
def _find_common_coerce(a, b):
if a > b:
return a
try:
thisind = __test_types.index(a.char)
except ValueError:
return None
while thisind < __len_test_types:
newdtype = dtype(__test_types[thisind])
if newdtype >= b and newdtype >= a:
return newdtype
thisind += 1
return None
def bitname(obj):
"""Return a bit-width name for a given type object"""
bits = _bits_of(obj)
char = dtype(obj).kind
base = _kind_to_stem[char]
if base == 'object':
bits = 0
if bits != 0:
char = "%s%d" % (char, bits // 8)
return base, bits, char
def _var(a, axis=None, dtype=None, out=None, ddof=0, keepdims=False):
arr = asanyarray(a)
rcount = _count_reduce_items(arr, axis)
# Make this warning show up on top.
if ddof >= rcount:
warnings.warn("Degrees of freedom <= 0 for slice", RuntimeWarning)
# Cast bool, unsigned int, and int to float64 by default
if dtype is None and issubclass(arr.dtype.type, (nt.integer, nt.bool_)):
dtype = mu.dtype('f8')
# Compute the mean.
# Note that if dtype is not of inexact type then arraymean will
# not be either.
arrmean = um.add.reduce(arr, axis=axis, dtype=dtype, keepdims=True)
if isinstance(arrmean, mu.ndarray):
arrmean = um.true_divide(arrmean,
rcount,
out=arrmean,
casting='unsafe',
subok=False)
else:
arrmean = arrmean.dtype.type(arrmean / rcount)
# Compute sum of squared deviations from mean
# Note that x may not be inexact and that we need it to be an array,
# not a scalar.
x = asanyarray(arr - arrmean)
if issubclass(arr.dtype.type, nt.complexfloating):
x = um.multiply(x, um.conjugate(x), out=x).real
else:
x = um.multiply(x, x, out=x)
ret = um.add.reduce(x, axis=axis, dtype=dtype, out=out, keepdims=keepdims)
# Compute degrees of freedom and make sure it is not negative.
rcount = max([rcount - ddof, 0])
# divide by degrees of freedom
if isinstance(ret, mu.ndarray):
ret = um.true_divide(ret,
rcount,
out=ret,
casting='unsafe',
subok=False)
elif hasattr(ret, 'dtype'):
ret = ret.dtype.type(ret / rcount)
else:
ret = ret / rcount
return ret
def bitname(obj):
"""Return a bit-width name for a given type object"""
bits = _bits_of(obj)
dt = dtype(obj)
char = dt.kind
base = _kind_name(dt)
if base == "object":
bits = 0
if bits != 0:
char = "%s%d" % (char, bits // 8)
return base, bits, char
def _can_coerce_all(dtypelist, start=0):
N = len(dtypelist)
if N == 0:
return None
if N == 1:
return dtypelist[0]
thisind = start
while thisind < __len_test_types:
newdtype = dtype(__test_types[thisind])
numcoerce = len([x for x in dtypelist if newdtype >= x])
if numcoerce == N:
return newdtype
thisind += 1
return None
def _can_coerce_all(dtypelist, start=0):
N = len(dtypelist)
if N == 0:
return None
if N == 1:
return dtypelist[0]
thisind = start
while thisind < __len_test_types:
newdtype = dtype(__test_types[thisind])
numcoerce = len([x for x in dtypelist if newdtype >= x])
if numcoerce == N:
return newdtype
thisind += 1
return None
def _var(a, axis=None, dtype=None, out=None, ddof=0, keepdims=False):
arr = asanyarray(a)
rcount = _count_reduce_items(arr, axis)
# Make this warning show up on top.
if ddof >= rcount:
warnings.warn("Degrees of freedom <= 0 for slice",
RuntimeWarning,
stacklevel=2)
# Cast bool, unsigned int, and int to float64 by default
if dtype is None and issubclass(arr.dtype.type, (nt.integer, nt.bool_)):
dtype = mu.dtype('f8')
# Compute the mean.
# Note that if dtype is not of inexact type then arraymean will
# not be either.
arrmean = umr_sum(arr, axis, dtype, keepdims=True)
if isinstance(arrmean, mu.ndarray):
arrmean = um.true_divide(arrmean,
rcount,
out=arrmean,
casting='unsafe',
subok=False)
else:
arrmean = arrmean.dtype.type(arrmean / rcount)
# Compute sum of squared deviations from mean
ret = 0
for i in nm.nditer(arr, order='F'):
if issubclass(arr.dtype.type, nt.complexfloating):
ret += (i - arrmean * um.conjugate(i - arrmean)).real
else:
ret += (i - arrmean)**2
# Compute degrees of freedom and make sure it is not negative.
rcount = max([rcount - ddof, 0])
# divide by degrees of freedom
if isinstance(ret, mu.ndarray):
ret = um.true_divide(ret,
rcount,
out=ret,
casting='unsafe',
subok=False)
elif hasattr(ret, 'dtype'):
ret = ret.dtype.type(ret / rcount)
else:
ret = ret / rcount
return ret
def sctype2char(sctype):
"""
Return the string representation of a scalar dtype.
Parameters
----------
sctype : scalar dtype or object
If a scalar dtype, the corresponding string character is
returned. If an object, `sctype2char` tries to infer its scalar type
and then return the corresponding string character.
Returns
-------
typechar : str
The string character corresponding to the scalar type.
Raises
------
ValueError
If `sctype` is an object for which the type can not be inferred.
See Also
--------
obj2sctype, issctype, issubsctype, mintypecode
Examples
--------
>>> for sctype in [np.int32, float, complex, np.string_, np.ndarray]:
... print(np.sctype2char(sctype))
l
d
D
S
O
>>> x = np.array([1., 2-1.j])
>>> np.sctype2char(x)
'D'
>>> np.sctype2char(list)
'O'
"""
sctype = obj2sctype(sctype)
if sctype is None:
raise ValueError("unrecognized type")
if sctype not in _concrete_types:
# for compatibility
raise KeyError(sctype)
return dtype(sctype).char
def sctype2char(sctype):
"""
Return the string representation of a scalar dtype.
Parameters
----------
sctype : scalar dtype or object
If a scalar dtype, the corresponding string character is
returned. If an object, `sctype2char` tries to infer its scalar type
and then return the corresponding string character.
Returns
-------
typechar : str
The string character corresponding to the scalar type.
Raises
------
ValueError
If `sctype` is an object for which the type can not be inferred.
See Also
--------
obj2sctype, issctype, issubsctype, mintypecode
Examples
--------
>>> for sctype in [np.int32, np.double, np.complex_, np.string_, np.ndarray]:
... print(np.sctype2char(sctype))
l # may vary
d
D
S
O
>>> x = np.array([1., 2-1.j])
>>> np.sctype2char(x)
'D'
>>> np.sctype2char(list)
'O'
"""
sctype = obj2sctype(sctype)
if sctype is None:
raise ValueError("unrecognized type")
if sctype not in _concrete_types:
# for compatibility
raise KeyError(sctype)
return dtype(sctype).char
def _mean(a, axis=None, dtype=None, out=None, keepdims=False, *, where=True):
arr = asanyarray(a)
is_float16_result = False
rcount = _count_reduce_items(arr, axis, keepdims=keepdims, where=where)
if rcount == 0 if where is True else umr_any(rcount == 0, axis=None):
warnings.warn("Mean of empty slice.", RuntimeWarning, stacklevel=2)
# Cast bool, unsigned int, and int to float64 by default
if dtype is None:
if issubclass(arr.dtype.type, (nt.integer, nt.bool_)):
dtype = mu.dtype('f8')
elif issubclass(arr.dtype.type, nt.float16):
dtype = mu.dtype('f4')
is_float16_result = True
ret = umr_sum(arr, axis, dtype, out, keepdims, where=where)
if isinstance(ret, mu.ndarray):
with _no_nep50_warning():
ret = um.true_divide(ret,
rcount,
out=ret,
casting='unsafe',
subok=False)
if is_float16_result and out is None:
ret = arr.dtype.type(ret)
elif hasattr(ret, 'dtype'):
if is_float16_result:
ret = arr.dtype.type(ret / rcount)
else:
ret = ret.dtype.type(ret / rcount)
else:
ret = ret / rcount
return ret
def maximum_sctype(t):
"""
Return the scalar type of highest precision of the same kind as the input.
Parameters
----------
t : dtype or dtype specifier
The input data type. This can be a `dtype` object or an object that
is convertible to a `dtype`.
Returns
-------
out : dtype
The highest precision data type of the same kind (`dtype.kind`) as `t`.
See Also
--------
obj2sctype, mintypecode, sctype2char
dtype
Examples
--------
>>> np.maximum_sctype(int)
<type 'numpy.int64'>
>>> np.maximum_sctype(np.uint8)
<type 'numpy.uint64'>
>>> np.maximum_sctype(complex)
<type 'numpy.complex192'>
>>> np.maximum_sctype(str)
<type 'numpy.string_'>
>>> np.maximum_sctype('i2')
<type 'numpy.int64'>
>>> np.maximum_sctype('f4')
<type 'numpy.float96'>
"""
g = obj2sctype(t)
if g is None:
return t
t = g
bits = _bits_of(t)
base = _kind_to_stem[dtype(t).kind]
if base in sctypes:
return sctypes[base][-1]
else:
return t
def maximum_sctype(t):
"""
Return the scalar type of highest precision of the same kind as the input.
Parameters
----------
t : dtype or dtype specifier
The input data type. This can be a `dtype` object or an object that
is convertible to a `dtype`.
Returns
-------
out : dtype
The highest precision data type of the same kind (`dtype.kind`) as `t`.
See Also
--------
obj2sctype, mintypecode, sctype2char
dtype
Examples
--------
>>> np.maximum_sctype(int)
<type 'numpy.int64'>
>>> np.maximum_sctype(np.uint8)
<type 'numpy.uint64'>
>>> np.maximum_sctype(complex)
<type 'numpy.complex192'>
>>> np.maximum_sctype(str)
<type 'numpy.string_'>
>>> np.maximum_sctype('i2')
<type 'numpy.int64'>
>>> np.maximum_sctype('f4')
<type 'numpy.float96'>
"""
g = obj2sctype(t)
if g is None:
return t
t = g
bits = _bits_of(t)
base = _kind_to_stem[dtype(t).kind]
if base in sctypes:
return sctypes[base][-1]
else:
return t
def maximum_sctype(t):
"""
Return the scalar type of highest precision of the same kind as the input.
Parameters
----------
t : dtype or dtype specifier
The input data type. This can be a `dtype` object or an object that
is convertible to a `dtype`.
Returns
-------
out : dtype
The highest precision data type of the same kind (`dtype.kind`) as `t`.
See Also
--------
obj2sctype, mintypecode, sctype2char
dtype
Examples
--------
>>> np.maximum_sctype(int)
<class 'numpy.int64'>
>>> np.maximum_sctype(np.uint8)
<class 'numpy.uint64'>
>>> np.maximum_sctype(complex)
<class 'numpy.complex256'> # may vary
>>> np.maximum_sctype(str)
<class 'numpy.str_'>
>>> np.maximum_sctype('i2')
<class 'numpy.int64'>
>>> np.maximum_sctype('f4')
<class 'numpy.float128'> # may vary
"""
g = obj2sctype(t)
if g is None:
return t
t = g
base = _kind_name(dtype(t))
if base in sctypes:
return sctypes[base][-1]
else:
return t
def _var(a, axis=None, dtype=None, out=None, ddof=0, keepdims=False):
arr = asanyarray(a)
rcount = _count_reduce_items(arr, axis)
# Make this warning show up on top.
if ddof >= rcount:
warnings.warn("Degrees of freedom <= 0 for slice", RuntimeWarning,
stacklevel=2)
# Cast bool, unsigned int, and int to float64 by default
if dtype is None and issubclass(arr.dtype.type, (nt.integer, nt.bool_)):
dtype = mu.dtype('f8')
# Compute the mean.
# Note that if dtype is not of inexact type then arraymean will
# not be either.
arrmean = umr_sum(arr, axis, dtype, keepdims=True)
if isinstance(arrmean, mu.ndarray):
arrmean = um.true_divide(
arrmean, rcount, out=arrmean, casting='unsafe', subok=False)
else:
arrmean = arrmean.dtype.type(arrmean / rcount)
# Compute sum of squared deviations from mean
# Note that x may not be inexact and that we need it to be an array,
# not a scalar.
x = asanyarray(arr - arrmean)
if issubclass(arr.dtype.type, (nt.floating, nt.integer)):
x = um.multiply(x, x, out=x)
else:
x = um.multiply(x, um.conjugate(x), out=x).real
ret = umr_sum(x, axis, dtype, out, keepdims)
# Compute degrees of freedom and make sure it is not negative.
rcount = max([rcount - ddof, 0])
# divide by degrees of freedom
if isinstance(ret, mu.ndarray):
ret = um.true_divide(
ret, rcount, out=ret, casting='unsafe', subok=False)
elif hasattr(ret, 'dtype'):
ret = ret.dtype.type(ret / rcount)
else:
ret = ret / rcount
return ret
def maximum_sctype(t):
"""
Return the scalar type of highest precision of the same kind as the input.
Parameters
----------
t : dtype or dtype specifier
The input data type. This can be a `dtype` object or an object that
is convertible to a `dtype`.
Returns
-------
out : dtype
The highest precision data type of the same kind (`dtype.kind`) as `t`.
See Also
--------
obj2sctype, mintypecode, sctype2char
dtype
Examples
--------
>>> np.maximum_sctype(int)
<class 'numpy.int64'>
>>> np.maximum_sctype(np.uint8)
<class 'numpy.uint64'>
>>> np.maximum_sctype(complex)
<class 'numpy.complex256'> # may vary
>>> np.maximum_sctype(str)
<class 'numpy.str_'>
>>> np.maximum_sctype('i2')
<class 'numpy.int64'>
>>> np.maximum_sctype('f4')
<class 'numpy.float128'> # may vary
"""
g = obj2sctype(t)
if g is None:
return t
t = g
base = _kind_name(dtype(t))
if base in sctypes:
return sctypes[base][-1]
else:
return t
def _getintp_ctype():
from .multiarray import dtype
val = _getintp_ctype.cache
if val is not None:
return val
char = dtype('p').char
import ctypes
if (char == 'i'):
val = ctypes.c_int
elif char == 'l':
val = ctypes.c_long
elif char == 'q':
val = ctypes.c_longlong
else:
val = ctypes.c_long
_getintp_ctype.cache = val
return val
def _getintp_ctype():
from .multiarray import dtype
val = _getintp_ctype.cache
if val is not None:
return val
char = dtype('p').char
import ctypes
if (char == 'i'):
val = ctypes.c_int
elif char == 'l':
val = ctypes.c_long
elif char == 'q':
val = ctypes.c_longlong
else:
val = ctypes.c_long
_getintp_ctype.cache = val
return val
def obj2sctype(rep, default=None):
try:
if issubclass(rep, generic):
return rep
except TypeError:
pass
if isinstance(rep, dtype):
return rep.type
if isinstance(rep, type):
return _python_type(rep)
if isinstance(rep, ndarray):
return rep.dtype.type
try:
res = dtype(rep)
except:
return default
return res.type
def obj2sctype(rep, default=None):
try:
if issubclass(rep, generic):
return rep
except TypeError:
pass
if isinstance(rep, dtype):
return rep.type
if isinstance(rep, type):
return _python_type(rep)
if isinstance(rep, ndarray):
return rep.dtype.type
try:
res = dtype(rep)
except:
return default
return res.type
def _mean(a, axis=None, dtype=None, out=None, keepdims=False):
arr = asanyarray(a)
rcount = _count_reduce_items(arr, axis)
# Make this warning show up first
if rcount == 0:
warnings.warn("Mean of empty slice.", RuntimeWarning)
# Cast bool, unsigned int, and int to float64 by default
if dtype is None and issubclass(arr.dtype.type, (nt.integer, nt.bool_)):
dtype = mu.dtype('f8')
ret = um.add.reduce(arr, axis=axis, dtype=dtype, out=out, keepdims=keepdims)
if isinstance(ret, mu.ndarray):
ret = um.true_divide(
ret, rcount, out=ret, casting='unsafe', subok=False)
else:
ret = ret.dtype.type(ret / rcount)
return ret
def _mean(a, axis=None, dtype=None, out=None, keepdims=False):
arr = asanyarray(a)
rcount = _count_reduce_items(arr, axis)
# Make this warning show up first
if rcount == 0:
warnings.warn("Mean of empty slice.", RuntimeWarning)
# Cast bool, unsigned int, and int to float64 by default
if dtype is None and issubclass(arr.dtype.type, (nt.integer, nt.bool_)):
dtype = mu.dtype('f8')
ret = um.add.reduce(arr, axis=axis, dtype=dtype, out=out, keepdims=keepdims)
if isinstance(ret, mu.ndarray):
ret = um.true_divide(
ret, rcount, out=ret, casting='unsafe', subok=False)
else:
ret = ret.dtype.type(ret / rcount)
return ret
from __future__ import print_function
from _partition import lib, ffi
from _parition_build import list_suff, list_type
from numpy.core.multiarray import dtype
_type_to_suff = dict(zip(list_type, list_suff))
_dtype_to_cffi_type = {dtype('int32'): 'npy_int',
dtype('int64'): 'npy_longlong',
dtype('uint32'): 'npy_uint',
dtype('uint64'): 'npy_ulonglong',
dtype('float64'): 'npy_double',
dtype('float32'): 'npy_float',
}
def _cffi_type(dtype_input):
return _dtype_to_cffi_type.get(dtype_input)
class IndexesOverAxis(object):
"""
Class for iterating over an array along one axis. Similar functionality is implemented in numpy.apply_along_axis.
>>> indexes = IndexesOverAxis((2,3,3), 1)
>>> list(indexes)
[(0, slice(None, None, None), 0), (0, slice(None, None, None), 1), (0, slice(None, None, None), 2), (1, slice(None, None, None), 0), (1, slice(None, None, None), 1), (1, slice(None, None, None), 2)]
>>> indexes = IndexesOverAxis((2,2,3), 2)
>>> list(indexes)
[(0, 0, slice(None, None, None)), (0, 1, slice(None, None, None)), (1, 0, slice(None, None, None)), (1, 1, slice(None, None, None))]
def _dtype_from_pep3118(spec, byteorder='@', is_subdtype=False):
from numpy.core.multiarray import dtype
fields = {}
offset = 0
explicit_name = False
this_explicit_name = False
common_alignment = 1
is_padding = False
last_offset = 0
dummy_name_index = [0]
def next_dummy_name():
dummy_name_index[0] += 1
def get_dummy_name():
while True:
name = 'f%d' % dummy_name_index[0]
if name not in fields:
return name
next_dummy_name()
# Parse spec
while spec:
value = None
# End of structure, bail out to upper level
if spec[0] == '}':
spec = spec[1:]
break
# Sub-arrays (1)
shape = None
if spec[0] == '(':
j = spec.index(')')
shape = tuple(map(int, spec[1:j].split(',')))
spec = spec[j + 1:]
# Byte order
if spec[0] in ('@', '=', '<', '>', '^', '!'):
byteorder = spec[0]
if byteorder == '!':
byteorder = '>'
spec = spec[1:]
# Byte order characters also control native vs. standard type sizes
if byteorder in ('@', '^'):
type_map = _pep3118_native_map
type_map_chars = _pep3118_native_typechars
else:
type_map = _pep3118_standard_map
type_map_chars = _pep3118_standard_typechars
# Item sizes
itemsize = 1
if spec[0].isdigit():
j = 1
for j in range(1, len(spec)):
if not spec[j].isdigit():
break
itemsize = int(spec[:j])
spec = spec[j:]
# Data types
is_padding = False
if spec[:2] == 'T{':
value, spec, align, next_byteorder = _dtype_from_pep3118(
spec[2:], byteorder=byteorder, is_subdtype=True)
elif spec[0] in type_map_chars:
next_byteorder = byteorder
if spec[0] == 'Z':
j = 2
else:
j = 1
typechar = spec[:j]
spec = spec[j:]
is_padding = (typechar == 'x')
dtypechar = type_map[typechar]
if dtypechar in 'USV':
dtypechar += '%d' % itemsize
itemsize = 1
numpy_byteorder = {'@': '=', '^': '='}.get(byteorder, byteorder)
value = dtype(numpy_byteorder + dtypechar)
align = value.alignment
else:
raise ValueError("Unknown PEP 3118 data type specifier %r" % spec)
#
# Native alignment may require padding
#
# Here we assume that the presence of a '@' character implicitly implies
# that the start of the array is *already* aligned.
#
extra_offset = 0
if byteorder == '@':
start_padding = (-offset) % align
intra_padding = (-value.itemsize) % align
offset += start_padding
if intra_padding != 0:
if itemsize > 1 or (shape is not None and _prod(shape) > 1):
# Inject internal padding to the end of the sub-item
value = _add_trailing_padding(value, intra_padding)
else:
# We can postpone the injection of internal padding,
# as the item appears at most once
extra_offset += intra_padding
# Update common alignment
common_alignment = (align * common_alignment /
_gcd(align, common_alignment))
# Convert itemsize to sub-array
if itemsize != 1:
value = dtype((value, (itemsize, )))
# Sub-arrays (2)
if shape is not None:
value = dtype((value, shape))
# Field name
this_explicit_name = False
if spec and spec.startswith(':'):
i = spec[1:].index(':') + 1
name = spec[1:i]
spec = spec[i + 1:]
explicit_name = True
this_explicit_name = True
else:
name = get_dummy_name()
if not is_padding or this_explicit_name:
if name in fields:
raise RuntimeError(
"Duplicate field name '%s' in PEP3118 format" % name)
fields[name] = (value, offset)
last_offset = offset
if not this_explicit_name:
next_dummy_name()
byteorder = next_byteorder
offset += value.itemsize
offset += extra_offset
# Check if this was a simple 1-item type
if len(list(fields.keys())) == 1 and not explicit_name and fields['f0'][1] == 0 \
and not is_subdtype:
ret = fields['f0'][0]
else:
ret = dtype(fields)
# Trailing padding must be explicitly added
padding = offset - ret.itemsize
if byteorder == '@':
padding += (-offset) % common_alignment
if is_padding and not this_explicit_name:
ret = _add_trailing_padding(ret, padding)
# Finished
if is_subdtype:
return ret, spec, common_alignment, byteorder
else:
return ret
def find_common_type(array_types, scalar_types):
"""
Determine common type following standard coercion rules.
Parameters
----------
array_types : sequence
A list of dtypes or dtype convertible objects representing arrays.
scalar_types : sequence
A list of dtypes or dtype convertible objects representing scalars.
Returns
-------
datatype : dtype
The common data type, which is the maximum of `array_types` ignoring
`scalar_types`, unless the maximum of `scalar_types` is of a
different kind (`dtype.kind`). If the kind is not understood, then
None is returned.
See Also
--------
dtype, common_type, can_cast, mintypecode
Examples
--------
>>> np.find_common_type([], [np.int64, np.float32, complex])
dtype('complex128')
>>> np.find_common_type([np.int64, np.float32], [])
dtype('float64')
The standard casting rules ensure that a scalar cannot up-cast an
array unless the scalar is of a fundamentally different kind of data
(i.e. under a different hierarchy in the data type hierarchy) then
the array:
>>> np.find_common_type([np.float32], [np.int64, np.float64])
dtype('float32')
Complex is of a different type, so it up-casts the float in the
`array_types` argument:
>>> np.find_common_type([np.float32], [complex])
dtype('complex128')
Type specifier strings are convertible to dtypes and can therefore
be used instead of dtypes:
>>> np.find_common_type(['f4', 'f4', 'i4'], ['c8'])
dtype('complex128')
"""
array_types = [dtype(x) for x in array_types]
scalar_types = [dtype(x) for x in scalar_types]
maxa = _can_coerce_all(array_types)
maxsc = _can_coerce_all(scalar_types)
if maxa is None:
return maxsc
if maxsc is None:
return maxa
try:
index_a = _kind_list.index(maxa.kind)
index_sc = _kind_list.index(maxsc.kind)
except ValueError:
return None
if index_sc > index_a:
return _find_common_coerce(maxsc, maxa)
else:
return maxa
def issubdtype(arg1, arg2):
"""
Returns True if first argument is a typecode lower/equal in type hierarchy.
Parameters
----------
arg1, arg2 : dtype_like
dtype or string representing a typecode.
Returns
-------
out : bool
See Also
--------
issubsctype, issubclass_
numpy.core.numerictypes : Overview of numpy type hierarchy.
Examples
--------
>>> np.issubdtype('S1', np.string_)
True
>>> np.issubdtype(np.float64, np.float32)
False
"""
if not issubclass_(arg1, generic):
arg1 = dtype(arg1).type
if not issubclass_(arg2, generic):
arg2_orig = arg2
arg2 = dtype(arg2).type
if not isinstance(arg2_orig, dtype):
# weird deprecated behaviour, that tried to infer np.floating from
# float, and similar less obvious things, such as np.generic from
# basestring
mro = arg2.mro()
arg2 = mro[1] if len(mro) > 1 else mro[0]
def type_repr(x):
""" Helper to produce clear error messages """
if not isinstance(x, type):
return repr(x)
elif issubclass(x, generic):
return "np.{}".format(x.__name__)
else:
return x.__name__
# 1.14, 2017-08-01
warnings.warn(
"Conversion of the second argument of issubdtype from `{raw}` "
"to `{abstract}` is deprecated. In future, it will be treated "
"as `{concrete} == np.dtype({raw}).type`.".format(
raw=type_repr(arg2_orig),
abstract=type_repr(arg2),
concrete=type_repr(dtype(arg2_orig).type),
),
FutureWarning,
stacklevel=2,
)
return issubclass(arg1, arg2)
def _scalar_type_key(typ):
"""A ``key`` function for `sorted`."""
dt = dtype(typ)
return (dt.kind.lower(), dt.itemsize)
from __future__ import print_function
from _partition import lib, ffi
from _parition_build import list_suff, list_type
from numpy.core.multiarray import dtype
_type_to_suff = dict(zip(list_type, list_suff))
_dtype_to_cffi_type = {
dtype('int32'): 'npy_int',
dtype('int64'): 'npy_longlong',
dtype('uint32'): 'npy_uint',
dtype('uint64'): 'npy_ulonglong',
dtype('float64'): 'npy_double',
dtype('float32'): 'npy_float',
}
def _cffi_type(dtype_input):
return _dtype_to_cffi_type.get(dtype_input)
class IndexesOverAxis(object):
"""
Class for iterating over an array along one axis. Similar functionality is implemented in numpy.apply_along_axis.
>>> indexes = IndexesOverAxis((2,3,3), 1)
>>> list(indexes)
[(0, slice(None, None, None), 0), (0, slice(None, None, None), 1), (0, slice(None, None, None), 2), (1, slice(None, None, None), 0), (1, slice(None, None, None), 1), (1, slice(None, None, None), 2)]
>>> indexes = IndexesOverAxis((2,2,3), 2)
>>> list(indexes)
def find_common_type(array_types, scalar_types):
"""
Determine common type following standard coercion rules.
Parameters
----------
array_types : sequence
A list of dtypes or dtype convertible objects representing arrays.
scalar_types : sequence
A list of dtypes or dtype convertible objects representing scalars.
Returns
-------
datatype : dtype
The common data type, which is the maximum of `array_types` ignoring
`scalar_types`, unless the maximum of `scalar_types` is of a
different kind (`dtype.kind`). If the kind is not understood, then
None is returned.
See Also
--------
dtype, common_type, can_cast, mintypecode
Examples
--------
>>> np.find_common_type([], [np.int64, np.float32, np.complex])
dtype('complex128')
>>> np.find_common_type([np.int64, np.float32], [])
dtype('float64')
The standard casting rules ensure that a scalar cannot up-cast an
array unless the scalar is of a fundamentally different kind of data
(i.e. under a different hierarchy in the data type hierarchy) then
the array:
>>> np.find_common_type([np.float32], [np.int64, np.float64])
dtype('float32')
Complex is of a different type, so it up-casts the float in the
`array_types` argument:
>>> np.find_common_type([np.float32], [np.complex])
dtype('complex128')
Type specifier strings are convertible to dtypes and can therefore
be used instead of dtypes:
>>> np.find_common_type(['f4', 'f4', 'i4'], ['c8'])
dtype('complex128')
"""
array_types = [dtype(x) for x in array_types]
scalar_types = [dtype(x) for x in scalar_types]
maxa = _can_coerce_all(array_types)
maxsc = _can_coerce_all(scalar_types)
if maxa is None:
return maxsc
if maxsc is None:
return maxa
try:
index_a = _kind_list.index(maxa.kind)
index_sc = _kind_list.index(maxsc.kind)
except ValueError:
return None
if index_sc > index_a:
return _find_common_coerce(maxsc, maxa)
else:
return maxa
def _dtype_from_pep3118(spec, byteorder='@', is_subdtype=False):
from numpy.core.multiarray import dtype
fields = {}
offset = 0
explicit_name = False
this_explicit_name = False
common_alignment = 1
is_padding = False
last_offset = 0
dummy_name_index = [0]
def next_dummy_name():
dummy_name_index[0] += 1
def get_dummy_name():
while True:
name = 'f%d' % dummy_name_index[0]
if name not in fields:
return name
next_dummy_name()
# Parse spec
while spec:
value = None
# End of structure, bail out to upper level
if spec[0] == '}':
spec = spec[1:]
break
# Sub-arrays (1)
shape = None
if spec[0] == '(':
j = spec.index(')')
shape = tuple(map(int, spec[1:j].split(',')))
spec = spec[j+1:]
# Byte order
if spec[0] in ('@', '=', '<', '>', '^', '!'):
byteorder = spec[0]
if byteorder == '!':
byteorder = '>'
spec = spec[1:]
# Byte order characters also control native vs. standard type sizes
if byteorder in ('@', '^'):
type_map = _pep3118_native_map
type_map_chars = _pep3118_native_typechars
else:
type_map = _pep3118_standard_map
type_map_chars = _pep3118_standard_typechars
# Item sizes
itemsize = 1
if spec[0].isdigit():
j = 1
for j in range(1, len(spec)):
if not spec[j].isdigit():
break
itemsize = int(spec[:j])
spec = spec[j:]
# Data types
is_padding = False
if spec[:2] == 'T{':
value, spec, align, next_byteorder = _dtype_from_pep3118(
spec[2:], byteorder=byteorder, is_subdtype=True)
elif spec[0] in type_map_chars:
next_byteorder = byteorder
if spec[0] == 'Z':
j = 2
else:
j = 1
typechar = spec[:j]
spec = spec[j:]
is_padding = (typechar == 'x')
dtypechar = type_map[typechar]
if dtypechar in 'USV':
dtypechar += '%d' % itemsize
itemsize = 1
numpy_byteorder = {'@': '=', '^': '='}.get(byteorder, byteorder)
value = dtype(numpy_byteorder + dtypechar)
align = value.alignment
else:
raise ValueError("Unknown PEP 3118 data type specifier %r" % spec)
#
# Native alignment may require padding
#
# Here we assume that the presence of a '@' character implicitly implies
# that the start of the array is *already* aligned.
#
extra_offset = 0
if byteorder == '@':
start_padding = (-offset) % align
intra_padding = (-value.itemsize) % align
offset += start_padding
if intra_padding != 0:
if itemsize > 1 or (shape is not None and _prod(shape) > 1):
# Inject internal padding to the end of the sub-item
value = _add_trailing_padding(value, intra_padding)
else:
# We can postpone the injection of internal padding,
# as the item appears at most once
extra_offset += intra_padding
# Update common alignment
common_alignment = (align*common_alignment
/ _gcd(align, common_alignment))
# Convert itemsize to sub-array
if itemsize != 1:
value = dtype((value, (itemsize,)))
# Sub-arrays (2)
if shape is not None:
value = dtype((value, shape))
# Field name
this_explicit_name = False
if spec and spec.startswith(':'):
i = spec[1:].index(':') + 1
name = spec[1:i]
spec = spec[i+1:]
explicit_name = True
this_explicit_name = True
else:
name = get_dummy_name()
if not is_padding or this_explicit_name:
if name in fields:
raise RuntimeError("Duplicate field name '%s' in PEP3118 format"
% name)
fields[name] = (value, offset)
last_offset = offset
if not this_explicit_name:
next_dummy_name()
byteorder = next_byteorder
offset += value.itemsize
offset += extra_offset
# Check if this was a simple 1-item type
if len(fields) == 1 and not explicit_name and fields['f0'][1] == 0 \
and not is_subdtype:
ret = fields['f0'][0]
else:
ret = dtype(fields)
# Trailing padding must be explicitly added
padding = offset - ret.itemsize
if byteorder == '@':
padding += (-offset) % common_alignment
if is_padding and not this_explicit_name:
ret = _add_trailing_padding(ret, padding)
# Finished
if is_subdtype:
return ret, spec, common_alignment, byteorder
else:
return ret
def _var(a, axis=None, dtype=None, out=None, ddof=0, keepdims=False):
arr = asanyarray(a)
rcount = _count_reduce_items(arr, axis)
# Make this warning show up on top.
if ddof >= rcount:
warnings.warn("Degrees of freedom <= 0 for slice",
RuntimeWarning,
stacklevel=2)
# Cast bool, unsigned int, and int to float64 by default
if dtype is None and issubclass(arr.dtype.type, (nt.integer, nt.bool_)):
dtype = mu.dtype('f8')
# Compute the mean.
# Note that if dtype is not of inexact type then arraymean will
# not be either.
arrmean = umr_sum(arr, axis, dtype, keepdims=True)
if isinstance(arrmean, mu.ndarray):
arrmean = um.true_divide(arrmean,
rcount,
out=arrmean,
casting='unsafe',
subok=False)
else:
arrmean = arrmean.dtype.type(arrmean / rcount)
# Compute sum of squared deviations from mean
# Note that x may not be inexact and that we need it to be an array,
# not a scalar.
x = asanyarray(arr - arrmean)
if issubclass(arr.dtype.type, (nt.floating, nt.integer)):
x = um.multiply(x, x, out=x)
# Fast-paths for built-in complex types
elif x.dtype in _complex_to_float:
xv = x.view(dtype=(_complex_to_float[x.dtype], (2, )))
um.multiply(xv, xv, out=xv)
x = um.add(xv[..., 0], xv[..., 1], out=x.real).real
# Most general case; includes handling object arrays containing imaginary
# numbers and complex types with non-native byteorder
else:
x = um.multiply(x, um.conjugate(x), out=x).real
ret = umr_sum(x, axis, dtype, out, keepdims)
# Compute degrees of freedom and make sure it is not negative.
rcount = max([rcount - ddof, 0])
# divide by degrees of freedom
if isinstance(ret, mu.ndarray):
ret = um.true_divide(ret,
rcount,
out=ret,
casting='unsafe',
subok=False)
elif hasattr(ret, 'dtype'):
ret = ret.dtype.type(ret / rcount)
else:
ret = ret / rcount
return ret
def _var(a,
axis=None,
dtype=None,
out=None,
ddof=0,
keepdims=False,
*,
where=True):
arr = asanyarray(a)
rcount = _count_reduce_items(arr, axis, keepdims=keepdims, where=where)
# Make this warning show up on top.
if ddof >= rcount if where is True else umr_any(ddof >= rcount, axis=None):
warnings.warn("Degrees of freedom <= 0 for slice",
RuntimeWarning,
stacklevel=2)
# Cast bool, unsigned int, and int to float64 by default
if dtype is None and issubclass(arr.dtype.type, (nt.integer, nt.bool_)):
dtype = mu.dtype("f8")
# Compute the mean.
# Note that if dtype is not of inexact type then arraymean will
# not be either.
arrmean = umr_sum(arr, axis, dtype, keepdims=True, where=where)
# The shape of rcount has to match arrmean to not change the shape of out
# in broadcasting. Otherwise, it cannot be stored back to arrmean.
if rcount.ndim == 0:
# fast-path for default case when where is True
div = rcount
else:
# matching rcount to arrmean when where is specified as array
div = rcount.reshape(arrmean.shape)
if isinstance(arrmean, mu.ndarray):
arrmean = um.true_divide(arrmean,
div,
out=arrmean,
casting="unsafe",
subok=False)
else:
arrmean = arrmean.dtype.type(arrmean / rcount)
# Compute sum of squared deviations from mean
# Note that x may not be inexact and that we need it to be an array,
# not a scalar.
x = asanyarray(arr - arrmean)
if issubclass(arr.dtype.type, (nt.floating, nt.integer)):
x = um.multiply(x, x, out=x)
# Fast-paths for built-in complex types
elif x.dtype in _complex_to_float:
xv = x.view(dtype=(_complex_to_float[x.dtype], (2, )))
um.multiply(xv, xv, out=xv)
x = um.add(xv[..., 0], xv[..., 1], out=x.real).real
# Most general case; includes handling object arrays containing imaginary
# numbers and complex types with non-native byteorder
else:
x = um.multiply(x, um.conjugate(x), out=x).real
ret = umr_sum(x, axis, dtype, out, keepdims=keepdims, where=where)
# Compute degrees of freedom and make sure it is not negative.
rcount = um.maximum(rcount - ddof, 0)
# divide by degrees of freedom
if isinstance(ret, mu.ndarray):
ret = um.true_divide(ret,
rcount,
out=ret,
casting="unsafe",
subok=False)
elif hasattr(ret, "dtype"):
ret = ret.dtype.type(ret / rcount)
else:
ret = ret / rcount
return ret
def issubdtype(arg1, arg2):
r"""
Returns True if first argument is a typecode lower/equal in type hierarchy.
This is like the builtin :func:`issubclass`, but for `dtype`\ s.
Parameters
----------
arg1, arg2 : dtype_like
`dtype` or object coercible to one
Returns
-------
out : bool
See Also
--------
:ref:`arrays.scalars` : Overview of the numpy type hierarchy.
issubsctype, issubclass_
Examples
--------
`issubdtype` can be used to check the type of arrays:
>>> ints = np.array([1, 2, 3], dtype=np.int32)
>>> np.issubdtype(ints.dtype, np.integer)
True
>>> np.issubdtype(ints.dtype, np.floating)
False
>>> floats = np.array([1, 2, 3], dtype=np.float32)
>>> np.issubdtype(floats.dtype, np.integer)
False
>>> np.issubdtype(floats.dtype, np.floating)
True
Similar types of different sizes are not subdtypes of each other:
>>> np.issubdtype(np.float64, np.float32)
False
>>> np.issubdtype(np.float32, np.float64)
False
but both are subtypes of `floating`:
>>> np.issubdtype(np.float64, np.floating)
True
>>> np.issubdtype(np.float32, np.floating)
True
For convenience, dtype-like objects are allowed too:
>>> np.issubdtype('S1', np.string_)
True
>>> np.issubdtype('i4', np.signedinteger)
True
"""
if not issubclass_(arg1, generic):
arg1 = dtype(arg1).type
if not issubclass_(arg2, generic):
arg2 = dtype(arg2).type
return issubclass(arg1, arg2)
# with the final arguments, no broadcasting needed.
import os, sys
from ._umath_linalg_build import all_four, three
from ._umath_linalg_cffi import ffi, lib
lib.init_constants()
import numpy as np
# dtype has not been imported yet. Fake it.
from numpy.core.multiarray import dtype
class Dummy(object):
pass
nt = Dummy()
nt.int32 = dtype('int32')
nt.int8 = dtype('int8')
nt.float32 = dtype('float32')
nt.float64 = dtype('float64')
nt.complex64 = dtype('complex64')
nt.complex128 = dtype('complex128')
from numpy.core.umath import frompyfunc
__version__ = '0.1.4'
def toCharP(src):
if src is None:
return ffi.cast('void*', 0)
pData = src.__array_interface__['data'][0]
return ffi.cast('char *', pData)
# TODO macos?
if sys.platform == 'win32':
so_name = '/umath_linalg_cffi.dll'
else:
so_name = '/libumath_linalg_cffi.so'
umath_linalg_capi = umath_ffi.dlopen(os.path.dirname(__file__) + so_name)
umath_linalg_capi.init_constants()
import numpy as np
# dtype has not been imported yet. Fake it.
from numpy.core.multiarray import dtype
class Dummy(object):
pass
nt = Dummy()
nt.int32 = dtype('int32')
nt.int8 = dtype('int8')
nt.float32 = dtype('float32')
nt.float64 = dtype('float64')
nt.complex64 = dtype('complex64')
nt.complex128 = dtype('complex128')
from numpy.core.umath import frompyfunc
__version__ = '0.1.4'
toCtypeA = {nt.int32: 'int[1]', nt.float32: 'float[1]', nt.float64: 'double[1]',
nt.complex64: 'f2c_complex[1]', nt.complex128: 'f2c_doublecomplex[1]'}
def toCharP(src):
if src is None:
return umath_ffi.cast('void*', 0)
