PyPyNum 1.18.0__py3-none-any.whl → 1.18.1__py3-none-any.whl

pypynum/README.md

@@ -19,7 +19,7 @@ science, and ML, it offers efficient, general-purpose tools.
 [![Downloads](https://static.pepy.tech/badge/pypynum/month)](https://pepy.tech/project/pypynum)
 [![Downloads](https://static.pepy.tech/badge/pypynum/week)](https://pepy.tech/project/pypynum)
 
-## PyPyNum | Version -> 1.18.0 | PyPI -> https://pypi.org/project/PyPyNum/ | Gitee -> https://www.gitee.com/PythonSJL/PyPyNum | GitHub -> https://github.com/PythonSJL/PyPyNum
+## PyPyNum | Version -> 1.18.1 | PyPI -> https://pypi.org/project/PyPyNum/ | Gitee -> https://www.gitee.com/PythonSJL/PyPyNum | GitHub -> https://github.com/PythonSJL/PyPyNum
 
 ![LOGO](PyPyNum.png)
 
@@ -148,70 +148,10 @@ Python interpreter and run it!
 
 ### Functional Changes Compared to the Previous Version
 
-```
-!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=
-
-PyPyNum version 1.18.0 has undergone the following functional changes compared to version 1.17.2:
-
-1. `arrays` Module
-   a. Fixed reflection operation errors in the `Array` class.
-
-2. `ciphers` Module
-   a. The `base_64` function's input parameter and return type have been changed from `str` to `bytes`.
-
-3. `groups` Module
-   a. The `Group` class has been refactored and new available methods have been added.
-   b. The `group` function now accepts an additional `operation` parameter.
-
-4. `images` Module
-   a. Added a new class `JPEG` for JPEG image handling.
-   b. Added JPEG processing functions: `jpeg_adjust_qtable`, `jpeg_category`, `jpeg_channel_encoding`,
-      `jpeg_chroma_dc_huff`, `jpeg_dct8x8`, `jpeg_decode_pixels`, `jpeg_encode_pixels`, `jpeg_luma_dc_huff`,
-      `jpeg_rle_decoding`, `jpeg_rle_encoding`, `jpeg_split_pixels`, and `jpeg_zigzag`.
-   c. Added color space conversion functions `rgb2ycbcr` and `ycbcr2rgb` (which are JPEG-related functions).
-   d. Added `entropy` function (which is a PNG function).
-   e. Added PNG filter functions `png_apply_filter` and `png_reverse_filter`, allowing the `apply_filter` parameter
-      to be specified during writing to compress the generated file size.
-
-5. `kernels` Module
-   a. Added `matmul8x8kernel` function for 8x8 matrix multiplication kernel operations.
-
-6. `maths` Module
-   a. Added `dsigmoid` function.
-   b. Updated `sumprod` function with a new `floating` parameter.
-
-7. `matrices` Module
-   a. Added `dctmtx` function to generate a discrete cosine transform matrix.
-
-8. `numbers` Module
-   a. Added `round_sigfig` function.
-   b. Added `words2int` function to convert English words to integers.
-
-9. `plotting` Module
-   a. Renamed the `color` function to `colortext`.
-
-10. `seqs` Module
-    a. Updated `stirling1` function with a new `sign` parameter.
-
-11. `special` Module
-    a. Added `ellipe` and `ellipk` functions for complete elliptic integrals.
-
-12. `symbols` Module
-    a. Added `Expr` class.
-    b. Added `build_expr_tree`, `infix2postfix`, and `tokenize` functions.
-    c. The `parse_expr` function now returns an `Expr` object (previously returned a `list`).
-
-13. `ufuncs` Module
-    a. Renamed comparison functions to follow more explicit naming conventions:
-       i. `eq` renamed to `equal`
-       ii. `ge` renamed to `greater_equal`
-       iii. `gt` renamed to `greater_than`
-       iv. `le` renamed to `less_equal`
-       v. `lt` renamed to `less_than`
-       vi. `ne` renamed to `not_equal`
-
-!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=!=
-```
+> 1. The `hypcmpnms` module added hypercomplex exponential and logarithmic operations.
+> 2. The `dists` module fixed all calculation errors.
+> 3. The `geoms` module fixed calculation errors and added more features.
+> 4. The `utils` module fixed calculation errors in the `IntervalSet` class.
 
 ### Run Time Test
 
@@ -239,23 +179,23 @@ Python interpreter version
 PyPyNum
 ├── arrays
 │   ├── CLASS
-│   │   ├── Array(object)/__init__(self: Any, data: Any, check: Any) -> Any
+│   │   ├── Array(object)/__init__(self: Any, data: Any, check: Any, dtype: Any) -> Any
 │   │   └── BoolArray(pypynum.arrays.Array)/__init__(self: Any, data: Any, check: Any) -> Any
 │   └── FUNCTION
-│       ├── array(data: Any) -> Any
-│       ├── asarray(data: Any) -> Any
+│       ├── array(data: Any, dtype: Any) -> Any
+│       ├── asarray(data: Any, dtype: Any) -> Any
 │       ├── aslist(data: Any) -> Any
 │       ├── boolarray(data: Any) -> Any
-│       ├── fill(shape: typing.Union[list, tuple], sequence: typing.Union[list, tuple, str], repeat: bool, pad: typing.Any, rtype: typing.Callable) -> typing.Any
-│       ├── full(shape: typing.Union[list, tuple], fill_value: typing.Any, rtype: typing.Callable) -> typing.Any
-│       ├── full_like(a: typing.Any, fill_value: typing.Any, rtype: typing.Callable) -> typing.Any
+│       ├── fill(shape: typing.Sequence, sequence: typing.Sequence, repeat: bool, pad: typing.Any, rtype: typing.Callable, dtype: type) -> typing.Any
+│       ├── full(shape: typing.Sequence, fill_value: typing.Any, rtype: typing.Callable, dtype: type) -> typing.Any
+│       ├── full_like(a: typing.Any, fill_value: typing.Any, rtype: typing.Callable, dtype: type) -> typing.Any
 │       ├── get_shape(data: Any) -> Any
 │       ├── is_valid_array(_array: Any, _shape: Any) -> Any
-│       ├── ones(shape: typing.Union[list, tuple], rtype: typing.Callable) -> typing.Any
-│       ├── ones_like(a: typing.Any, rtype: typing.Callable) -> typing.Any
+│       ├── ones(shape: typing.Sequence, rtype: typing.Callable, dtype: type) -> typing.Any
+│       ├── ones_like(a: typing.Any, rtype: typing.Callable, dtype: type) -> typing.Any
 │       ├── tensorproduct(tensors: pypynum.arrays.Array) -> pypynum.arrays.Array
-│       ├── zeros(shape: typing.Union[list, tuple], rtype: typing.Callable) -> typing.Any
-│       └── zeros_like(a: typing.Any, rtype: typing.Callable) -> typing.Any
+│       ├── zeros(shape: typing.Sequence, rtype: typing.Callable, dtype: type) -> typing.Any
+│       └── zeros_like(a: typing.Any, rtype: typing.Callable, dtype: type) -> typing.Any
 ├── chars
 │   ├── CLASS
 │   └── FUNCTION
@@ -297,7 +237,7 @@ PyPyNum
 ├── dists
 │   ├── CLASS
 │   └── FUNCTION
-│       ├── beta_pdf(x: Any, a: Any, b: Any) -> Any
+│       ├── beta_pdf(x: Any, alpha: Any, beta: Any) -> Any
 │       ├── binom_pmf(k: Any, n: Any, p: Any) -> Any
 │       ├── cauchy_cdf(x: Any, x0: Any, gamma: Any) -> Any
 │       ├── cauchy_pdf(x: Any, x0: Any, gamma: Any) -> Any
@@ -308,22 +248,22 @@ PyPyNum
 │       ├── f_pdf(x: Any, dfnum: Any, dfden: Any) -> Any
 │       ├── gamma_pdf(x: Any, shape: Any, scale: Any) -> Any
 │       ├── geometric_pmf(k: Any, p: Any) -> Any
-│       ├── hypergeom_pmf(k: Any, mg: Any, n: Any, nt: Any) -> Any
-│       ├── invgauss_pdf(x: Any, mu: Any, lambda_: Any, alpha: Any) -> Any
+│       ├── hypergeom_pmf(M: Any, n: Any, N: Any, k: Any) -> Any
+│       ├── invgauss_pdf(x: Any, mu: Any, lam: Any) -> Any
 │       ├── levy_pdf(x: Any, c: Any) -> Any
 │       ├── log_logistic_cdf(x: Any, alpha: Any, beta: Any) -> Any
 │       ├── log_logistic_pdf(x: Any, alpha: Any, beta: Any) -> Any
-│       ├── logistic_cdf(x: Any, mu: Any, s: Any) -> Any
-│       ├── logistic_pdf(x: Any, mu: Any, s: Any) -> Any
+│       ├── logistic_cdf(x: Any, loc: Any, scale: Any) -> Any
+│       ├── logistic_pdf(x: Any, loc: Any, scale: Any) -> Any
 │       ├── lognorm_cdf(x: Any, mu: Any, sigma: Any) -> Any
-│       ├── lognorm_pdf(x: Any, s: Any, scale: Any) -> Any
+│       ├── lognorm_pdf(x: Any, mu: Any, sigma: Any) -> Any
 │       ├── logser_pmf(k: Any, p: Any) -> Any
 │       ├── multinomial_pmf(k: Any, n: Any, p: Any) -> Any
 │       ├── nbinom_pmf(k: Any, n: Any, p: Any) -> Any
-│       ├── nhypergeom_pmf(k: Any, m: Any, n: Any, r: Any) -> Any
+│       ├── nhypergeom_pmf(M: Any, n: Any, r: Any, k: Any) -> Any
 │       ├── normal_cdf(x: Any, mu: Any, sigma: Any) -> Any
 │       ├── normal_pdf(x: Any, mu: Any, sigma: Any) -> Any
-│       ├── pareto_pdf(x: Any, k: Any, m: Any) -> Any
+│       ├── pareto_pdf(x: Any, b: Any, scale: Any) -> Any
 │       ├── poisson_pmf(k: Any, mu: Any) -> Any
 │       ├── rayleigh_pdf(x: Any, sigma: Any) -> Any
 │       ├── t_pdf(x: Any, df: Any) -> Any
@@ -356,6 +296,8 @@ PyPyNum
 │   │   ├── Quadrilateral(object)/__init__(self: Any, a: typing.Union[list, tuple], b: typing.Union[list, tuple], c: typing.Union[list, tuple], d: typing.Union[list, tuple]) -> Any
 │   │   └── Triangle(object)/__init__(self: Any, a: typing.Union[list, tuple], b: typing.Union[list, tuple], c: typing.Union[list, tuple]) -> Any
 │   └── FUNCTION
+│       ├── _rotate_point(p: Any, theta: Any, cx: Any, cy: Any) -> Any
+│       ├── _scale_point(p: Any, k: Any, cx: Any, cy: Any) -> Any
 │       └── distance(g1: Any, g2: Any, error: typing.Union[int, float]) -> float
 ├── graphs
 │   ├── CLASS
@@ -546,23 +488,23 @@ PyPyNum
 │       └── zeta(alpha: typing.Union[int, float, complex]) -> typing.Union[int, float, complex]
 ├── matrices
 │   ├── CLASS
-│   │   └── Matrix(pypynum.arrays.Array)/__init__(self: Any, data: Any, check: Any) -> Any
+│   │   └── Matrix(pypynum.arrays.Array)/__init__(self: Any, data: Any, check: Any, dtype: Any) -> Any
 │   └── FUNCTION
-│       ├── cholesky(matrix: pypynum.matrices.Matrix, hermitian: bool) -> pypynum.matrices.Matrix
-│       ├── dctmtx(n: int) -> pypynum.matrices.Matrix
-│       ├── diag(v: typing.Any, k: int, n: int, m: int) -> typing.Any
+│       ├── cholesky(matrix: pypynum.matrices.Matrix, hermitian: bool, dtype: Any) -> pypynum.matrices.Matrix
+│       ├── dctmtx(n: int, dtype: Any) -> pypynum.matrices.Matrix
+│       ├── diag(v: typing.Any, k: int, n: int, m: int, dtype: Any) -> typing.Any
 │       ├── diag_indices(n: int, k: int, m: int) -> tuple
-│       ├── eigen(matrix: pypynum.matrices.Matrix) -> tuple
-│       ├── hessenberg(matrix: pypynum.matrices.Matrix) -> tuple
-│       ├── identity(n: int, m: int) -> pypynum.matrices.Matrix
-│       ├── lu(matrix: pypynum.matrices.Matrix) -> tuple
-│       ├── mat(data: typing.Any) -> pypynum.matrices.Matrix
-│       ├── perm_mat(num_rows: int, num_cols: int, row_swaps: typing.Union[list, tuple], col_swaps: typing.Union[list, tuple], rtype: typing.Callable) -> typing.Any
+│       ├── eigen(matrix: pypynum.matrices.Matrix, dtype: Any) -> tuple
+│       ├── hessenberg(matrix: pypynum.matrices.Matrix, dtype: Any) -> tuple
+│       ├── identity(n: int, m: int, dtype: Any) -> pypynum.matrices.Matrix
+│       ├── lu(matrix: pypynum.matrices.Matrix, dtype: Any) -> tuple
+│       ├── mat(data: typing.Any, dtype: Any) -> pypynum.matrices.Matrix
+│       ├── perm_mat(num_rows: int, num_cols: int, row_swaps: typing.Union[list, tuple], col_swaps: typing.Union[list, tuple], rtype: typing.Callable, dtype: Any) -> typing.Any
 │       ├── perm_mat_indices(num_rows: int, num_cols: int, row_swaps: typing.Union[list, tuple], col_swaps: typing.Union[list, tuple]) -> tuple
-│       ├── qr(matrix: pypynum.matrices.Matrix, reduce: bool) -> tuple
-│       ├── rank_decomp(matrix: pypynum.matrices.Matrix) -> tuple
-│       ├── rotate90(matrix: pypynum.matrices.Matrix, times: int) -> pypynum.matrices.Matrix
-│       ├── svd(matrix: pypynum.matrices.Matrix, full: bool, calc_uv: bool) -> tuple
+│       ├── qr(matrix: pypynum.matrices.Matrix, reduce: bool, dtype: Any) -> tuple
+│       ├── rank_decomp(matrix: pypynum.matrices.Matrix, dtype: Any) -> tuple
+│       ├── rotate90(matrix: pypynum.matrices.Matrix, times: int, dtype: Any) -> pypynum.matrices.Matrix
+│       ├── svd(matrix: pypynum.matrices.Matrix, full: bool, calc_uv: bool, dtype: Any) -> tuple
 │       ├── tril_indices(n: int, k: int, m: int) -> tuple
 │       └── triu_indices(n: int, k: int, m: int) -> tuple
 ├── multiprec
@@ -801,7 +743,7 @@ PyPyNum
 │   └── FUNCTION
 ├── vectors
 │   ├── CLASS
-│   │   └── Vector(pypynum.arrays.Array)/__init__(self: Any, data: Any, check: Any) -> Any
+│   │   └── Vector(pypynum.arrays.Array)/__init__(self: Any, data: Any, check: Any, dtype: Any) -> Any
 │   └── FUNCTION
 │       └── vec(data: Any) -> Any
 └── zh_cn
@@ -1066,7 +1008,7 @@ print("Value of exp(e^2):", exp_e_squared)
 """
 Modulus of the complex1: 3.064156670102011971734464
 Square root of the complex2: 1.630869996965458756791024890326755654659+0.9631646481434200567938527724670041364524i
-Power of complex1 raised to complex2: 0.144632311554759549910373681031-0.17337039786352194417298867399i
+Power of complex1 raised to complex2: 0.144632311554759549910373681032-0.17337039786352194417298867399i
 Value of Euler's gamma constant: 0.577215664901532860606512090082402431042159336
 Logarithm of 2 (base 10): 0.30102999566398119521373889472449302676818988146211
 Value of exp(e^2): 1618.1779919126535017
@@ -1084,7 +1026,7 @@ Bessel Function of the first kind, order 0 at x=1: 0.7651976865579666
 Modified Bessel function of the first kind, order 1 at x=1: 0.565159103992485
 Hypergeometric function 0F1 at z=0.5 with b0=1: 1.5660829297563503
 Hypergeometric function 1F1 at z=1 with a0=1, b0=1: 2.7182818284590455
-q-Pochhammer Symbol with a=2+1j, q=0.5+0.1j, n=2+1j: (-0.33353429405776575+1.8573191887407854j)
+q-Pochhammer Symbol with a=2+1j, q=0.5+0.1j, n=2+1j: (-0.3335342940577659+1.8573191887407858j)
 q-Gamma Function at n=2 with q=0.5+0.1j: (0.9999999999999989-4.953510875709033e-16j)
 """
 
@@ -1161,7 +1103,7 @@ print(equations.poly_eq(p))
 print(equations.lin_eq(*m))
 
 """
-[(-1.5615528128088307-6.5209667308287455e-24j), (1.0000000000000007+3.241554513744382e-25j), (2.5615528128088294+4.456233626665941e-24j)]
+[(-1.5615528128088307-3.728101026218476e-24j), (1.0000000000000007+3.2415545137443745e-25j), (2.5615528128088294+7.352235277305158e-24j)]
 [1.6666666666666665, -0.6666666666666666, -0.4444444444444444]
 """
 
@@ -1262,10 +1204,10 @@ print(random.randint(0, 9, [2, 3, 4]))
 print(random.uniform(0, 9, [2, 3, 4]))
 
 """
-[[[0.19564108603837504, 1.01920370166293, -0.0432622197313817, 1.8767392228517594], [-0.9372140720220945, 0.6455939848661518, 0.06219926062283103, -0.03646501719525719], [-0.6575861228704779, -1.627831653990843, 1.4669927522358261, -2.1082184047042944]], [[-0.4550416369776341, 0.5921202822490813, 0.1393176939653395, -0.1994214479986489], [-0.559257919413219, -1.629764620332396, 0.04658277611512766, 0.20529751771180757], [0.492656546311569, 0.9731757491363898, -0.3281597719355017, -0.6855996837842998]]]
-[[[0.9897608447208633, 0.9091539554606014, 0.839742901348473, 0.7483640950392159], [0.1271954637618754, 0.23914774960655272, 0.6603167703557995, 0.09706202212722614], [0.1020987366479954, 0.16628877700140854, 0.9844464087410161, 0.8213963537473721]], [[0.268434775377716, 0.8021623176316018, 0.48319517625501174, 0.30410149780944096], [0.8970313978138931, 0.9980790184879383, 0.1185467392465952, 0.19570961919567464], [0.08626090792820118, 0.5790211560246683, 0.7475290741004703, 0.8048613204213184]]]
-[[[9, 0, 7, 5], [2, 7, 3, 5], [7, 6, 4, 2]], [[3, 4, 2, 0], [4, 2, 2, 1], [4, 5, 4, 9]]]
-[[[8.067902792666303, 3.1398970785364244, 7.875028237778611, 7.766921948037462], [3.337337214716076, 2.852783099370506, 1.028257336777862, 7.822605535085232], [7.111199916783942, 5.890725764603359, 3.038605829888671, 6.272148674191553]], [[1.076715403659712, 6.806086377982102, 5.699002266942416, 1.1654651363361368], [1.3920577362733448, 8.527352563852686, 2.359964996027981, 5.183071574749471], [7.8250627862940725, 1.6259461468536924, 8.333173856724002, 5.068032713313034]]]
+[[[0.18059388030329554, -0.7950286594701896, -1.6124456807502878, 1.0650032690802622], [1.231592625021916, -0.2923569307790834, 0.5743684840351643, -0.5994993318165575], [-0.4541501878112403, 0.5194701735167844, -0.41134655581914514, 2.5838345235765336]], [[-0.38027869224951916, 1.407233237769525, -1.4230691454968791, -1.098291804499891], [0.0720517812951513, 1.2716687933757764, -0.389868632414012, 0.18416602669163057], [-0.49978319823217315, 0.8789409340735714, -1.4538419291500941, -1.895514116697364]]]
+[[[0.3214199347664204, 0.4682526051010606, 0.8691333897517276, 0.65768740769836], [0.7632355843244432, 0.5804412886619936, 0.008332295847854443, 0.6151292129754637], [0.9795343167986345, 0.5205461670247327, 0.7248923173697739, 0.11717223436961488]], [[0.7662005170926139, 0.9714376279553422, 0.9550672285743181, 0.08793525427569937], [0.6339393968200449, 0.8301210945735973, 0.6564657233275701, 0.6220565155685059], [0.21406744938064437, 0.3047415909582225, 0.6510510008125593, 0.12269523561892148]]]
+[[[5, 1, 4, 9], [9, 1, 6, 2], [4, 3, 1, 9]], [[0, 9, 6, 9], [7, 8, 0, 2], [9, 8, 5, 3]]]
+[[[1.6943948141992893, 2.247908784078809, 8.414928873268128, 4.628393016930337], [7.572639917918929, 5.357050653981085, 8.240417270343459, 3.24879148260883], [4.771181663822661, 1.905700180078211, 3.7868896315863747, 4.2241419541638265]], [[4.635584514037156, 7.401849961502754, 0.6758543569780475, 6.101106332926675], [1.9978290038252968, 7.441511948419005, 0.10317848285562836, 5.723737686439019], [0.13415224754500865, 7.69176143777073, 2.9628506743472967, 5.077060383033804]]]
 """
 
 print(regs.lin_reg(list(range(5)), [2, 4, 6, 7, 8]))
@@ -1284,7 +1226,7 @@ print(tools.frange(0, 3, 0.4))
 print(tools.linspace(0, 2.8, 8))
 
 """
-{<class 'int'>: [1], <class 'float'>: [2.3, 3.14], <class 'complex'>: [(4+5j)], <class 'str'>: ['string'], <class 'type'>: [<class 'list'>, <class 'tuple'>], <class 'bool'>: [True, False], <class 'module'>: [<module 'pypynum.tools' from 'C:\\Users\\Administrator\\PycharmProjects\\pythonProject\\pypynum\\tools.py'>]}
+{<class 'int'>: [1], <class 'float'>: [2.3, 3.14], <class 'complex'>: [(4+5j)], <class 'str'>: ['string'], <class 'type'>: [<class 'list'>, <class 'tuple'>], <class 'bool'>: [True, False], <class 'module'>: [<module 'pypynum.tools' from 'C:\\Users\\Administrator\\PycharmProjects\\PythonProject\\pypynum\\tools.py'>]}
 ['Python', 6, 'NumPy', <class 'int'>, 'PyPyNum', 9, 'pypynum', True]
 [0.0, 0.4, 0.8, 1.2000000000000002, 1.6, 2.0, 2.4000000000000004, 2.8000000000000003, 3.2]
 [0.0, 0.39999999999999997, 0.7999999999999999, 1.2, 1.5999999999999999, 1.9999999999999998, 2.4, 2.8]

pypynum/__init__.py

@@ -60,7 +60,7 @@ from .seqs import *
 from .special import *
 from .stattest import *
 from .symbols import *
-from .tensors import *
+from . import tensors
 from .tools import *
 from .trees import *
 from .types import config
@@ -68,7 +68,7 @@ from .ufuncs import *
 from .utils import *
 from .vectors import *
 
-__version__ = "1.18.0"
+__version__ = "1.18.1"
 print("PyPyNum", "Version -> " + __version__, "PyPI -> https://pypi.org/project/PyPyNum/",
       "Gitee -> https://www.gitee.com/PythonSJL/PyPyNum", "GitHub -> https://github.com/PythonSJL/PyPyNum", sep=" | ")
 for key, value in tuple(globals().items()):

pypynum/arrays.py

@@ -1,4 +1,4 @@
-from .types import Any, Callable, ShapeError, arr, ite
+from .types import Any, Callable, Sequence, ShapeError
 
 ArrayError = ShapeError("The shape of the array is invalid")
 MatchError = ShapeError("The shapes of the two arrays do not match")
@@ -9,11 +9,31 @@ class Array:
     It is the base class of vectors and matrices, supporting operations and many statistical functions.
     :param data: An array in the form of a list
     :param check: Check the rationality of the input array
+    :param dtype: The data type of the elements in the array, default is object
     """
 
-    def __init__(self, data=None, check=True):
+    def __init__(self, data=None, check=True, dtype=object):
         if data is None:
             data = []
+        self.dtype = dtype
+
+        def _convert_list(data, dtype):
+            if isinstance(data, list):
+                return [_convert_list(item, dtype) for item in data]
+            else:
+                try:
+                    return dtype(data)
+                except (ValueError, TypeError):
+                    return data
+
+        if dtype is not object:
+            if isinstance(data, list):
+                data = _convert_list(data, dtype)
+            else:
+                try:
+                    data = dtype(data)
+                except (ValueError, TypeError):
+                    pass
         self.shape = [] if data == [] else get_shape(data)
         if check and self.shape and not isinstance(data, (int, float, complex)):
             is_valid_array(data, self.shape)
@@ -102,7 +122,8 @@ class Array:
         if isinstance(other, Array):
             if self.shape != other.shape:
                 raise MatchError
-            return fill(self.shape, [t1 + t2 for t1, t2 in zip(self.flatten(), other.flatten())], rtype=type(self))
+            return fill(self.shape, [t1 + t2 for t1, t2 in zip(self.flatten(), other.flatten())], rtype=type(self),
+                        dtype=self.dtype)
         elif isinstance(other, (int, float, complex)):
             from .ufuncs import add
             return add(self, other)
@@ -113,7 +134,8 @@ class Array:
         if isinstance(other, Array):
             if self.shape != other.shape:
                 raise MatchError
-            return fill(self.shape, [t1 - t2 for t1, t2 in zip(self.flatten(), other.flatten())], rtype=type(self))
+            return fill(self.shape, [t1 - t2 for t1, t2 in zip(self.flatten(), other.flatten())], rtype=type(self),
+                        dtype=self.dtype)
         elif isinstance(other, (int, float, complex)):
             from .ufuncs import subtract
             return subtract(self, other)
@@ -124,7 +146,8 @@ class Array:
         if isinstance(other, Array):
             if self.shape != other.shape:
                 raise MatchError
-            return fill(self.shape, [t1 * t2 for t1, t2 in zip(self.flatten(), other.flatten())], rtype=type(self))
+            return fill(self.shape, [t1 * t2 for t1, t2 in zip(self.flatten(), other.flatten())], rtype=type(self),
+                        dtype=self.dtype)
         elif isinstance(other, (int, float, complex)):
             from .ufuncs import multiply
             return multiply(self, other)
@@ -135,7 +158,8 @@ class Array:
         if isinstance(other, Array):
             if self.shape != other.shape:
                 raise MatchError
-            return fill(self.shape, [t1 / t2 for t1, t2 in zip(self.flatten(), other.flatten())], rtype=type(self))
+            return fill(self.shape, [t1 / t2 for t1, t2 in zip(self.flatten(), other.flatten())], rtype=type(self),
+                        dtype=self.dtype)
         elif isinstance(other, (int, float, complex)):
             from .ufuncs import divide
             return divide(self, other)
@@ -146,7 +170,8 @@ class Array:
         if isinstance(other, Array):
             if self.shape != other.shape:
                 raise MatchError
-            return fill(self.shape, [t1 // t2 for t1, t2 in zip(self.flatten(), other.flatten())], rtype=type(self))
+            return fill(self.shape, [t1 // t2 for t1, t2 in zip(self.flatten(), other.flatten())], rtype=type(self),
+                        dtype=self.dtype)
         elif isinstance(other, (int, float, complex)):
             from .ufuncs import floor_divide
             return floor_divide(self, other)
@@ -157,7 +182,8 @@ class Array:
         if isinstance(other, Array):
             if self.shape != other.shape:
                 raise MatchError
-            return fill(self.shape, [t1 % t2 for t1, t2 in zip(self.flatten(), other.flatten())], rtype=type(self))
+            return fill(self.shape, [t1 % t2 for t1, t2 in zip(self.flatten(), other.flatten())], rtype=type(self),
+                        dtype=self.dtype)
         elif isinstance(other, (int, float, complex)):
             from .ufuncs import modulo
             return modulo(self, other)
@@ -177,31 +203,31 @@ class Array:
 
     def __rsub__(self, other):
         if isinstance(other, (int, float, complex)):
-            return fill(self.shape, [other - t for t in self.flatten()], rtype=type(self))
+            return fill(self.shape, [other - t for t in self.flatten()], rtype=type(self), dtype=self.dtype)
         else:
             raise ValueError("Another must be an array or number")
 
     def __rtruediv__(self, other):
         if isinstance(other, (int, float, complex)):
-            return fill(self.shape, [other / t for t in self.flatten()], rtype=type(self))
+            return fill(self.shape, [other / t for t in self.flatten()], rtype=type(self), dtype=self.dtype)
         else:
             raise ValueError("Another must be an array or number")
 
     def __rfloordiv__(self, other):
         if isinstance(other, (int, float, complex)):
-            return fill(self.shape, [other // t for t in self.flatten()], rtype=type(self))
+            return fill(self.shape, [other // t for t in self.flatten()], rtype=type(self), dtype=self.dtype)
         else:
             raise ValueError("Another must be an array or number")
 
     def __rmod__(self, other):
         if isinstance(other, (int, float, complex)):
-            return fill(self.shape, [other % t for t in self.flatten()], rtype=type(self))
+            return fill(self.shape, [other % t for t in self.flatten()], rtype=type(self), dtype=self.dtype)
         else:
             raise ValueError("Another must be an array or number")
 
     def __rpow__(self, other):
         if isinstance(other, (int, float, complex)):
-            return fill(self.shape, [other ** t for t in self.flatten()], rtype=type(self))
+            return fill(self.shape, [other ** t for t in self.flatten()], rtype=type(self), dtype=self.dtype)
         else:
             raise ValueError("Exponential must be an array or number")
 
@@ -245,7 +271,7 @@ class Array:
         return data
 
     def reshape(self, shape, repeat=True, pad=0):
-        return fill(shape, self.flatten(), repeat, pad, type(self))
+        return fill(shape, self.flatten(), repeat, pad, type(self), dtype=self.dtype)
 
     def copy(self):
         from copy import deepcopy
@@ -263,7 +289,7 @@ class Array:
             else:
                 return [inner(list(_), 0) for _ in zip(*a)]
 
-        return type(self)(inner(self, axis))
+        return type(self)(inner(self, axis), dtype=self.dtype)
 
     def sum(self, axis=None):
         return self.basic(sum, axis)
@@ -337,8 +363,8 @@ def is_valid_array(_array, _shape):
     inner(_array, _shape, len(_array), shape_len)
 
 
-def array(data=None):
-    return Array(data)
+def array(data=None, dtype=object):
+    return Array(data, dtype=dtype)
 
 
 def aslist(data):
@@ -353,49 +379,65 @@ def aslist(data):
     raise TypeError("Unable to convert to array type")
 
 
-def asarray(data):
-    return Array(aslist(data))
+def asarray(data, dtype=object):
+    return Array(aslist(data), dtype=dtype)
 
 
-def full(shape: arr, fill_value: Any, rtype: Callable = Array) -> Any:
+def full(shape: Sequence, fill_value: Any, rtype: Callable = Array, dtype: type = object) -> Any:
     def inner(data, length):
         return fill_value if length == 0 else [inner(data[1:], length - 1) for _ in range(data[0])]
 
     if isinstance(fill_value, list):
         raise TypeError("The filled value cannot be a list")
+    if dtype is not object:
+        try:
+            fill_value = dtype(fill_value)
+        except (ValueError, TypeError):
+            pass
     result = inner(shape, len(shape))
-    return result if rtype is list else rtype(result)
+    return result if rtype is list else rtype(result, dtype=dtype)
 
 
-def full_like(a: Any, fill_value: Any, rtype: Callable = Array) -> Any:
+def full_like(a: Any, fill_value: Any, rtype: Callable = Array, dtype: type = None) -> Any:
     def inner(data):
         return [inner(item) for item in data] if isinstance(data, list) else fill_value
 
     if isinstance(fill_value, list):
         raise TypeError("The filled value cannot be a list")
+    if dtype is None:
+        if isinstance(a, Array):
+            dtype = a.dtype
+        else:
+            dtype = object
     if isinstance(a, Array):
         a = a.data
+    if dtype is not object:
+        try:
+            fill_value = dtype(fill_value)
+        except (ValueError, TypeError):
+            pass
     result = inner(a)
-    return result if rtype is list else rtype(result)
+    return result if rtype is list else rtype(result, dtype=dtype)
 
 
-def zeros(shape: arr, rtype: Callable = Array) -> Any:
-    return full(shape, 0, rtype)
+def zeros(shape: Sequence, rtype: Callable = Array, dtype: type = object) -> Any:
+    return full(shape, 0, rtype, dtype)
 
 
-def zeros_like(a: Any, rtype: Callable = Array) -> Any:
-    return full_like(a, 0, rtype)
+def zeros_like(a: Any, rtype: Callable = Array, dtype: type = None) -> Any:
+    return full_like(a, 0, rtype, dtype)
 
 
-def ones(shape: arr, rtype: Callable = Array) -> Any:
-    return full(shape, 1, rtype)
+def ones(shape: Sequence, rtype: Callable = Array, dtype: type = object) -> Any:
+    return full(shape, 1, rtype, dtype)
 
 
-def ones_like(a: Any, rtype: Callable = Array) -> Any:
-    return full_like(a, 1, rtype)
+def ones_like(a: Any, rtype: Callable = Array, dtype: type = None) -> Any:
+    return full_like(a, 1, rtype, dtype)
 
 
-def fill(shape: arr, sequence: ite = None, repeat: bool = True, pad: Any = 0, rtype: Callable = Array) -> Any:
+def fill(shape: Sequence, sequence: Sequence = None, repeat: bool = True, pad: Any = 0, rtype: Callable = Array,
+         dtype: type = object) -> Any:
     pointer = -1
     length = 1
     for item in shape:
@@ -416,7 +458,9 @@ def fill(shape: arr, sequence: ite = None, repeat: bool = True, pad: Any = 0, rt
             return [inner(_shape[1:], depth - 1) for _ in range(_shape[0])]
 
     result = inner(shape, len(shape))
-    return result if rtype is list else rtype(result)
+    if rtype is list:
+        return result
+    return rtype(result, dtype=dtype)
 
 
 def tensorproduct(*tensors: Array) -> Array:
@@ -428,7 +472,9 @@ def tensorproduct(*tensors: Array) -> Array:
     def mul(a, b):
         flattened_a = a.flatten()
         flattened_b = b.flatten()
-        return Array(fill(a.shape + b.shape, [i * j for i in flattened_a for j in flattened_b], rtype=list), False)
+        res_dtype = a.dtype if a.dtype is not object else b.dtype
+        data = fill(a.shape + b.shape, [i * j for i in flattened_a for j in flattened_b], rtype=list)
+        return Array(data, check=False, dtype=res_dtype)
 
     first = tensors[0]
     for second in tensors[1:]:
@@ -438,8 +484,7 @@ def tensorproduct(*tensors: Array) -> Array:
 
 class BoolArray(Array):
     def __init__(self, data=None, check=True):
-        from .ufuncs import apply
-        super().__init__(apply(Array(data, check), bool, list), False)
+        super().__init__(data, check, dtype=bool)
 
     def logic_op(self, other, func):
         from .ufuncs import ufunc_helper