37 KiB
37 KiB
Всё в Python является объектом¶
In [1]:
print(isinstance("add", object)) print(isinstance(1_000, object)) print(isinstance(3.14, object))
True True True
In [3]:
(3.14).as_integer_ratio()
Out[3]:
(7070651414971679, 2251799813685248)
In [13]:
class Vector2D: x = 0 y = 0 def norm(self): return (self.x**2 + self.y**2)**0.5 vec = Vector2D()
In [4]:
isinstance(vec, object)
Out[4]:
True
In [5]:
isinstance(Vector2D, object)
Out[5]:
True
In [4]:
isinstance(object, object)
Out[4]:
True
In [5]:
import math isinstance(math, object)
Out[5]:
True
In [6]:
def add(a,b): return a + b isinstance(add, object)
Out[6]:
True
In [7]:
add.x = 1
Магические методы¶
In [11]:
dir(vec)
Out[11]:
['__class__', '__delattr__', '__dict__', '__dir__', '__doc__', '__eq__', '__format__', '__ge__', '__getattribute__', '__gt__', '__hash__', '__init__', '__init_subclass__', '__le__', '__lt__', '__module__', '__ne__', '__new__', '__reduce__', '__reduce_ex__', '__repr__', '__setattr__', '__sizeof__', '__str__', '__subclasshook__', '__weakref__', 'norm', 'x', 'y']
In [18]:
dir(add)
Out[18]:
['__annotations__', '__call__', '__class__', '__closure__', '__code__', '__defaults__', '__delattr__', '__dict__', '__dir__', '__doc__', '__eq__', '__format__', '__ge__', '__get__', '__getattribute__', '__globals__', '__gt__', '__hash__', '__init__', '__init_subclass__', '__kwdefaults__', '__le__', '__lt__', '__module__', '__name__', '__ne__', '__new__', '__qualname__', '__reduce__', '__reduce_ex__', '__repr__', '__setattr__', '__sizeof__', '__str__', '__subclasshook__']
- Управляют внутренней работой объектов
- Хранят различную информацию объектов (которую можно получать в runtime)
- Вызываются при использовании синтаксических конструкций
- Вызываются встроенными (builtins) функциями
- Область применения: перегрузка операторов, рефлексия и метапрограммирование
In [8]:
class TenItemList: def __len__(self): return 10 ten_item_list = TenItemList() len(ten_item_list)
Out[8]:
10
Всё в Python является объектом, а все синтаксические конструкции сводятся к вызовам магических методов¶
Пример сложение¶
In [10]:
class Vector2D: def __init__(self, x, y): self.x = x self.y = y def __add__(self, other): return Vector2D(self.x + other.y, self.x + other.y) def norm(self): return (self.x**2 + self.y**2)**0.5 vec1 = Vector2D(1,2) vec2 = Vector2D(3,4) vec3 = vec1 + vec2 vec3.x, vec3.y
Out[10]:
(5, 5)
Пример присваивание¶
In [14]:
class Vector2D: x = 0 y = 0 def norm(self): return (self.x**2 + self.y**2)**0.5 vec = Vector2D()
In [15]:
vec = Vector2D() vec.__getattribute__("x")
Out[15]:
0
In [17]:
vec.__getattribute__("norm")()
Out[17]:
0.0
In [18]:
vec.x = 5 vec.__getattribute__("x")
Out[18]:
5
In [19]:
vec.__setattr__("x", 10) getattr(vec, "x")
Out[19]:
10
In [20]:
setattr(vec, "x", 20) vec.x
Out[20]:
20
In [21]:
class Foo: def __setattr__(self, key, value): print(key, value) foo = Foo() foo.a = "A"
a A
In [22]:
foo.a
--------------------------------------------------------------------------- AttributeError Traceback (most recent call last) /tmp/ipykernel_35789/2615815247.py in <module> ----> 1 foo.a AttributeError: 'Foo' object has no attribute 'a'
На самом деле все объекты реализованы как словари хранящие атрибуты объекта (однако есть возможности для оптимизаций)¶
In [23]:
class Vector2D: x = 0 y = 0 def norm(self): return (self.x**2 + self.y**2)**0.5 vec = Vector2D()
In [24]:
vec.__dict__
Out[24]:
{}
In [27]:
Vector2D.__dict__
Out[27]:
mappingproxy({'__module__': '__main__',
'x': 0,
'y': 0,
'norm': <function __main__.Vector2D.norm(self)>,
'__dict__': <attribute '__dict__' of 'Vector2D' objects>,
'__weakref__': <attribute '__weakref__' of 'Vector2D' objects>,
'__doc__': None})
In [26]:
vec.x = 5 vec.__dict__
Out[26]:
{'x': 5}
Модуль inspect --- информация об объектах в runtime¶
- Не вся информация может быть доступна через магические методы
- Недоступную информацию можно получить через модуль inspect
In [28]:
import inspect def add(a,b): return a + b inspect.isfunction(add)
Out[28]:
True
In [29]:
inspect.getsource(add)
Out[29]:
'def add(a,b): return a + b\n'
In [30]:
from numpy import random inspect.getsource(random)
Out[30]:
'"""\n========================\nRandom Number Generation\n========================\n\nUse ``default_rng()`` to create a `Generator` and call its methods.\n\n=============== =========================================================\nGenerator\n--------------- ---------------------------------------------------------\nGenerator Class implementing all of the random number distributions\ndefault_rng Default constructor for ``Generator``\n=============== =========================================================\n\n============================================= ===\nBitGenerator Streams that work with Generator\n--------------------------------------------- ---\nMT19937\nPCG64\nPCG64DXSM\nPhilox\nSFC64\n============================================= ===\n\n============================================= ===\nGetting entropy to initialize a BitGenerator\n--------------------------------------------- ---\nSeedSequence\n============================================= ===\n\n\nLegacy\n------\n\nFor backwards compatibility with previous versions of numpy before 1.17, the\nvarious aliases to the global `RandomState` methods are left alone and do not\nuse the new `Generator` API.\n\n==================== =========================================================\nUtility functions\n-------------------- ---------------------------------------------------------\nrandom Uniformly distributed floats over ``[0, 1)``\nbytes Uniformly distributed random bytes.\npermutation Randomly permute a sequence / generate a random sequence.\nshuffle Randomly permute a sequence in place.\nchoice Random sample from 1-D array.\n==================== =========================================================\n\n==================== =========================================================\nCompatibility\nfunctions - removed\nin the new API\n-------------------- ---------------------------------------------------------\nrand Uniformly distributed values.\nrandn Normally distributed values.\nranf Uniformly distributed floating point numbers.\nrandom_integers Uniformly distributed integers in a given range.\n (deprecated, use ``integers(..., closed=True)`` instead)\nrandom_sample Alias for `random_sample`\nrandint Uniformly distributed integers in a given range\nseed Seed the legacy random number generator.\n==================== =========================================================\n\n==================== =========================================================\nUnivariate\ndistributions\n-------------------- ---------------------------------------------------------\nbeta Beta distribution over ``[0, 1]``.\nbinomial Binomial distribution.\nchisquare :math:`\\\\chi^2` distribution.\nexponential Exponential distribution.\nf F (Fisher-Snedecor) distribution.\ngamma Gamma distribution.\ngeometric Geometric distribution.\ngumbel Gumbel distribution.\nhypergeometric Hypergeometric distribution.\nlaplace Laplace distribution.\nlogistic Logistic distribution.\nlognormal Log-normal distribution.\nlogseries Logarithmic series distribution.\nnegative_binomial Negative binomial distribution.\nnoncentral_chisquare Non-central chi-square distribution.\nnoncentral_f Non-central F distribution.\nnormal Normal / Gaussian distribution.\npareto Pareto distribution.\npoisson Poisson distribution.\npower Power distribution.\nrayleigh Rayleigh distribution.\ntriangular Triangular distribution.\nuniform Uniform distribution.\nvonmises Von Mises circular distribution.\nwald Wald (inverse Gaussian) distribution.\nweibull Weibull distribution.\nzipf Zipf\'s distribution over ranked data.\n==================== =========================================================\n\n==================== ==========================================================\nMultivariate\ndistributions\n-------------------- ----------------------------------------------------------\ndirichlet Multivariate generalization of Beta distribution.\nmultinomial Multivariate generalization of the binomial distribution.\nmultivariate_normal Multivariate generalization of the normal distribution.\n==================== ==========================================================\n\n==================== =========================================================\nStandard\ndistributions\n-------------------- ---------------------------------------------------------\nstandard_cauchy Standard Cauchy-Lorentz distribution.\nstandard_exponential Standard exponential distribution.\nstandard_gamma Standard Gamma distribution.\nstandard_normal Standard normal distribution.\nstandard_t Standard Student\'s t-distribution.\n==================== =========================================================\n\n==================== =========================================================\nInternal functions\n-------------------- ---------------------------------------------------------\nget_state Get tuple representing internal state of generator.\nset_state Set state of generator.\n==================== =========================================================\n\n\n"""\n__all__ = [\n \'beta\',\n \'binomial\',\n \'bytes\',\n \'chisquare\',\n \'choice\',\n \'dirichlet\',\n \'exponential\',\n \'f\',\n \'gamma\',\n \'geometric\',\n \'get_state\',\n \'gumbel\',\n \'hypergeometric\',\n \'laplace\',\n \'logistic\',\n \'lognormal\',\n \'logseries\',\n \'multinomial\',\n \'multivariate_normal\',\n \'negative_binomial\',\n \'noncentral_chisquare\',\n \'noncentral_f\',\n \'normal\',\n \'pareto\',\n \'permutation\',\n \'poisson\',\n \'power\',\n \'rand\',\n \'randint\',\n \'randn\',\n \'random\',\n \'random_integers\',\n \'random_sample\',\n \'ranf\',\n \'rayleigh\',\n \'sample\',\n \'seed\',\n \'set_state\',\n \'shuffle\',\n \'standard_cauchy\',\n \'standard_exponential\',\n \'standard_gamma\',\n \'standard_normal\',\n \'standard_t\',\n \'triangular\',\n \'uniform\',\n \'vonmises\',\n \'wald\',\n \'weibull\',\n \'zipf\',\n]\n\n# add these for module-freeze analysis (like PyInstaller)\nfrom . import _pickle\nfrom . import _common\nfrom . import _bounded_integers\n\nfrom ._generator import Generator, default_rng\nfrom .bit_generator import SeedSequence, BitGenerator\nfrom ._mt19937 import MT19937\nfrom ._pcg64 import PCG64, PCG64DXSM\nfrom ._philox import Philox\nfrom ._sfc64 import SFC64\nfrom .mtrand import *\n\n__all__ += [\'Generator\', \'RandomState\', \'SeedSequence\', \'MT19937\',\n \'Philox\', \'PCG64\', \'PCG64DXSM\', \'SFC64\', \'default_rng\',\n \'BitGenerator\']\n\n\ndef __RandomState_ctor():\n """Return a RandomState instance.\n\n This function exists solely to assist (un)pickling.\n\n Note that the state of the RandomState returned here is irrelevant, as this\n function\'s entire purpose is to return a newly allocated RandomState whose\n state pickle can set. Consequently the RandomState returned by this function\n is a freshly allocated copy with a seed=0.\n\n See https://github.com/numpy/numpy/issues/4763 for a detailed discussion\n\n """\n return RandomState(seed=0)\n\n\nfrom numpy._pytesttester import PytestTester\ntest = PytestTester(__name__)\ndel PytestTester\n'
Модуль inspect --- информация об объектах в runtime¶
- Не вся информация может быть доступна через магические методы
- Недоступную информацию можно получить через модуль inspect
In [2]:
import inspect def add(a,b): return a + b inspect.isfunction(add)
Out[2]:
True
In [12]:
inspect.getsource(add)
Out[12]:
'def add(a,b): return a + b\n'
In [3]:
from numpy import random inspect.getsource(random)
Out[3]:
'"""\n========================\nRandom Number Generation\n========================\n\nUse ``default_rng()`` to create a `Generator` and call its methods.\n\n=============== =========================================================\nGenerator\n--------------- ---------------------------------------------------------\nGenerator Class implementing all of the random number distributions\ndefault_rng Default constructor for ``Generator``\n=============== =========================================================\n\n============================================= ===\nBitGenerator Streams that work with Generator\n--------------------------------------------- ---\nMT19937\nPCG64\nPCG64DXSM\nPhilox\nSFC64\n============================================= ===\n\n============================================= ===\nGetting entropy to initialize a BitGenerator\n--------------------------------------------- ---\nSeedSequence\n============================================= ===\n\n\nLegacy\n------\n\nFor backwards compatibility with previous versions of numpy before 1.17, the\nvarious aliases to the global `RandomState` methods are left alone and do not\nuse the new `Generator` API.\n\n==================== =========================================================\nUtility functions\n-------------------- ---------------------------------------------------------\nrandom Uniformly distributed floats over ``[0, 1)``\nbytes Uniformly distributed random bytes.\npermutation Randomly permute a sequence / generate a random sequence.\nshuffle Randomly permute a sequence in place.\nchoice Random sample from 1-D array.\n==================== =========================================================\n\n==================== =========================================================\nCompatibility\nfunctions - removed\nin the new API\n-------------------- ---------------------------------------------------------\nrand Uniformly distributed values.\nrandn Normally distributed values.\nranf Uniformly distributed floating point numbers.\nrandom_integers Uniformly distributed integers in a given range.\n (deprecated, use ``integers(..., closed=True)`` instead)\nrandom_sample Alias for `random_sample`\nrandint Uniformly distributed integers in a given range\nseed Seed the legacy random number generator.\n==================== =========================================================\n\n==================== =========================================================\nUnivariate\ndistributions\n-------------------- ---------------------------------------------------------\nbeta Beta distribution over ``[0, 1]``.\nbinomial Binomial distribution.\nchisquare :math:`\\\\chi^2` distribution.\nexponential Exponential distribution.\nf F (Fisher-Snedecor) distribution.\ngamma Gamma distribution.\ngeometric Geometric distribution.\ngumbel Gumbel distribution.\nhypergeometric Hypergeometric distribution.\nlaplace Laplace distribution.\nlogistic Logistic distribution.\nlognormal Log-normal distribution.\nlogseries Logarithmic series distribution.\nnegative_binomial Negative binomial distribution.\nnoncentral_chisquare Non-central chi-square distribution.\nnoncentral_f Non-central F distribution.\nnormal Normal / Gaussian distribution.\npareto Pareto distribution.\npoisson Poisson distribution.\npower Power distribution.\nrayleigh Rayleigh distribution.\ntriangular Triangular distribution.\nuniform Uniform distribution.\nvonmises Von Mises circular distribution.\nwald Wald (inverse Gaussian) distribution.\nweibull Weibull distribution.\nzipf Zipf\'s distribution over ranked data.\n==================== =========================================================\n\n==================== ==========================================================\nMultivariate\ndistributions\n-------------------- ----------------------------------------------------------\ndirichlet Multivariate generalization of Beta distribution.\nmultinomial Multivariate generalization of the binomial distribution.\nmultivariate_normal Multivariate generalization of the normal distribution.\n==================== ==========================================================\n\n==================== =========================================================\nStandard\ndistributions\n-------------------- ---------------------------------------------------------\nstandard_cauchy Standard Cauchy-Lorentz distribution.\nstandard_exponential Standard exponential distribution.\nstandard_gamma Standard Gamma distribution.\nstandard_normal Standard normal distribution.\nstandard_t Standard Student\'s t-distribution.\n==================== =========================================================\n\n==================== =========================================================\nInternal functions\n-------------------- ---------------------------------------------------------\nget_state Get tuple representing internal state of generator.\nset_state Set state of generator.\n==================== =========================================================\n\n\n"""\n__all__ = [\n \'beta\',\n \'binomial\',\n \'bytes\',\n \'chisquare\',\n \'choice\',\n \'dirichlet\',\n \'exponential\',\n \'f\',\n \'gamma\',\n \'geometric\',\n \'get_state\',\n \'gumbel\',\n \'hypergeometric\',\n \'laplace\',\n \'logistic\',\n \'lognormal\',\n \'logseries\',\n \'multinomial\',\n \'multivariate_normal\',\n \'negative_binomial\',\n \'noncentral_chisquare\',\n \'noncentral_f\',\n \'normal\',\n \'pareto\',\n \'permutation\',\n \'poisson\',\n \'power\',\n \'rand\',\n \'randint\',\n \'randn\',\n \'random\',\n \'random_integers\',\n \'random_sample\',\n \'ranf\',\n \'rayleigh\',\n \'sample\',\n \'seed\',\n \'set_state\',\n \'shuffle\',\n \'standard_cauchy\',\n \'standard_exponential\',\n \'standard_gamma\',\n \'standard_normal\',\n \'standard_t\',\n \'triangular\',\n \'uniform\',\n \'vonmises\',\n \'wald\',\n \'weibull\',\n \'zipf\',\n]\n\n# add these for module-freeze analysis (like PyInstaller)\nfrom . import _pickle\nfrom . import _common\nfrom . import _bounded_integers\n\nfrom ._generator import Generator, default_rng\nfrom .bit_generator import SeedSequence, BitGenerator\nfrom ._mt19937 import MT19937\nfrom ._pcg64 import PCG64, PCG64DXSM\nfrom ._philox import Philox\nfrom ._sfc64 import SFC64\nfrom .mtrand import *\n\n__all__ += [\'Generator\', \'RandomState\', \'SeedSequence\', \'MT19937\',\n \'Philox\', \'PCG64\', \'PCG64DXSM\', \'SFC64\', \'default_rng\',\n \'BitGenerator\']\n\n\ndef __RandomState_ctor():\n """Return a RandomState instance.\n\n This function exists solely to assist (un)pickling.\n\n Note that the state of the RandomState returned here is irrelevant, as this\n function\'s entire purpose is to return a newly allocated RandomState whose\n state pickle can set. Consequently the RandomState returned by this function\n is a freshly allocated copy with a seed=0.\n\n See https://github.com/numpy/numpy/issues/4763 for a detailed discussion\n\n """\n return RandomState(seed=0)\n\n\nfrom numpy._pytesttester import PytestTester\ntest = PytestTester(__name__)\ndel PytestTester\n'
In [ ]: