NumPy 1.14.2官方文档：数组操作与通用函数详解

需积分: 9 45 浏览量更新于2024-07-18 收藏 7.81MB PDF 举报

本文档是关于NumPy库的1.14-2版本的帮助文档，NumPy是Python中用于科学计算的核心库，提供高性能的多维数组对象和大量的数学函数。这个版本包含了关键的API方法和使用示例，让开发者能更好地理解和利用该库。 1. **Array objects** (一维及多维数组): NumPy的核心是其ndarray对象，它是N维数组，可以处理各种数据类型，包括整数、浮点数、复数以及字符串。ndarray提供了高效的内存管理，支持广播规则，使得简单的元素级操作变得容易且高效。 2. **Datatypes and dtype objects** (数据类型): 数据类型（dtype）用于描述数组中的元素类型，如int64、float32等。了解并正确选择合适的dtype对于性能和内存效率至关重要。 3. **Indexing** (索引和切片): 文档详细介绍了如何通过索引和切片访问数组的不同部分，这对于数据操作和数据分析非常重要。 4. **Iterating over arrays** (遍历数组): 提供了对数组元素进行迭代的方法，这在处理数组数据时必不可少，例如循环或逐元素处理。 5. **Standard array subclasses** (标准数组子类): 可能包括稀疏数组、矩阵等特殊类型的数组，这些子类扩展了基本ndarray的功能。 6. **Masked arrays** (掩码数组): 用于处理缺失值或不完整数据的数组，这对于科学计算中的异常值处理非常有用。 7. **The Array Interface** (数组接口): 这是NumPy与其他语言和库交互的标准，确保了与其他支持此接口的库（如SciPy）的兼容性。 8. **Datetime and Timedelta support** (日期时间处理): 提供了针对日期和时间数据的特定功能，包括格式转换和算术运算。 9. **Universal functions (ufuncs)**: 通用函数是NumPy的灵魂，它们是高阶的数学函数，可以对数组进行元素级操作，支持广播和类型转换。 10. **Routines** (内置函数): 分类涵盖了创建数组、数组操作、二进制运算、字符串操作、C-类型接口、日期时间支持函数、数据类型特定函数、SciPy加速的函数、数学函数、浮点错误处理和离散傅里叶变换等多个方面的功能。 11. **Optional SciPy-accelerated routines** 和 **Mathematical functions with automatic domain**: 部分函数可能利用了SciPy库的优化，提供更高级的数学计算功能，而无需用户显式调用SciPy。通过阅读这篇文档，开发者可以全面掌握NumPy 1.14-2版的使用，无论是对数组操作有深入理解，还是想要提高代码的性能，都能从中找到所需的信息。同时，它也是学习NumPy的最佳参考资料，适用于从初学者到高级开发者的不同层次。

NumPy Reference, Release 1.14.2

This array is stored in memory as 40 bytes, one after the other (known as a contiguous block of memory).

The strides of an array tell us how many bytes we have to skip in memory to move to the next position

along a certain axis. For example, we have to skip 4 bytes (1 value) to move to the next column, but 20

bytes (5 values) to get to the same position in the next row. As such, the strides for the array x will be

(20, 4).

Examples

>>> y = np.reshape(np.arange(2

4), (2,3,4))

>>> y

array([[[ 0, 1, 2, 3],

[ 4, 5, 6, 7],

[ 8, 9, 10, 11]],

[[12, 13, 14, 15],

[16, 17, 18, 19],

[20, 21, 22, 23]]])

>>> y.strides

(48, 16, 4)

>>> y[1,1,1]

>>> offset=sum(y.strides

np.array((1,1,1)))

>>> offset/y.itemsize

>>> x = np.reshape(np.arange(5

8), (5,6,7,8)).transpose(2,3,1,0)

>>> x.strides

(32, 4, 224, 1344)

>>> i = np.array([3,5,2,2])

>>> offset = sum(i

x.strides)

>>> x[3,5,2,2]

813

>>> offset / x.itemsize

813

ndarray.ctypes

An object to simplify the interaction of the array with the ctypes module.

This attribute creates an object that makes it easier to use arrays when calling shared libraries with the

ctypes module. The returned object has, among others, data, shape, and strides attributes (see Notes

below) which themselves return ctypes objects that can be used as arguments to a shared library.

Parameters None

Returns c : Python object

Possessing attributes data, shape, strides, etc.

See also:

numpy.ctypeslib

Notes

Below are the public attributes of this object which were documented in “Guide to NumPy” (we have

omitted undocumented public attributes, as well as documented private attributes):

12 Chapter 1. Array objects

NumPy Reference, Release 1.14.2

• data: A pointer to the memory area of the array as a Python integer. This memory area may contain

data that is not aligned, or not in correct byte-order. The memory area may not even be writeable.

The array ﬂags and data-type of this array should be respected when passing this attribute to arbitrary

C-code to avoid trouble that can include Python crashing. User Beware! The value of this attribute is

exactly the same as self._array_interface_[‘data’][0].

• shape (c_intp*self.ndim): A ctypes array of length self.ndim where the basetype is the C-integer

corresponding to dtype(‘p’) on this platform. This base-type could be c_int, c_long, or c_longlong

depending on the platform. The c_intp type is deﬁned accordingly in numpy.ctypeslib. The ctypes

array contains the shape of the underlying array.

• strides (c_intp*self.ndim): A ctypes array of length self.ndim where the basetype is the same as for

the shape attribute. This ctypes array contains the strides information from the underlying array.

This strides information is important for showing how many bytes must be jumped to get to the next

element in the array.

• data_as(obj): Return the data pointer cast to a particular c-types object. For example, calling

self._as_parameter_ is equivalent to self.data_as(ctypes.c_void_p). Perhaps you want to use the data

as a pointer to a ctypes array of ﬂoating-point data: self.data_as(ctypes.POINTER(ctypes.c_double)).

• shape_as(obj): Return the shape tuple as an array of some other c-types type. For example:

self.shape_as(ctypes.c_short).

• strides_as(obj): Return the strides tuple as an array of some other c-types type. For example:

self.strides_as(ctypes.c_longlong).

Be careful using the ctypes attribute - especially on temporary arrays or arrays constructed on the ﬂy. For

example, calling (a+b).ctypes.data_as(ctypes.c_void_p) returns a pointer to memory that

is invalid because the array created as (a+b) is deallocated before the next Python statement. You can avoid

this problem using either c=a+b or ct=(a+b).ctypes. In the latter case, ct will hold a reference to

the array until ct is deleted or re-assigned.

If the ctypes module is not available, then the ctypes attribute of array objects still returns something useful,

but ctypes objects are not returned and errors may be raised instead. In particular, the object will still have

the as parameter attribute which will return an integer equal to the data attribute.

Examples

>>> import ctypes

>>> x

array([[0, 1],

[2, 3]])

>>> x.ctypes.data

30439712

>>> x.ctypes.data_as(ctypes.POINTER(ctypes.c_long))

<ctypes.LP_c_long object at 0x01F01300>

>>> x.ctypes.data_as(ctypes.POINTER(ctypes.c_long)).contents

c_long(0)

>>> x.ctypes.data_as(ctypes.POINTER(ctypes.c_longlong)).contents

c_longlong(4294967296L)

>>> x.ctypes.shape

<numpy.core._internal.c_long_Array_2 object at 0x01FFD580>

>>> x.ctypes.shape_as(ctypes.c_long)

<numpy.core._internal.c_long_Array_2 object at 0x01FCE620>

>>> x.ctypes.strides

<numpy.core._internal.c_long_Array_2 object at 0x01FCE620>

>>> x.ctypes.strides_as(ctypes.c_longlong)

<numpy.core._internal.c_longlong_Array_2 object at 0x01F01300>

1.1. The N-dimensional array (ndarray) 13

NumPy Reference, Release 1.14.2

Table 1.3 – continued from previous page

item(*args) Copy an element of an array to a standard Python scalar

and return it.

itemset(*args) Insert scalar into an array (scalar is cast to array’s dtype,

if possible)

max([axis, out, keepdims]) Return the maximum along a given axis.

mean([axis, dtype, out, keepdims]) Returns the average of the array elements along given

axis.

min([axis, out, keepdims]) Return the minimum along a given axis.

newbyteorder([new_order]) Return the array with the same data viewed with a dif-

ferent byte order.

nonzero() Return the indices of the elements that are non-zero.

partition(kth[, axis, kind, order]) Rearranges the elements in the array in such a way that

value of the element in kth position is in the position it

would be in a sorted array.

prod([axis, dtype, out, keepdims]) Return the product of the array elements over the given

axis

ptp([axis, out]) Peak to peak (maximum - minimum) value along a

given axis.

put(indices, values[, mode]) Set a.flat[n] = values[n] for all n in indices.

ravel([order]) Return a ﬂattened array.

repeat(repeats[, axis]) Repeat elements of an array.

reshape(shape[, order]) Returns an array containing the same data with a new

shape.

resize(new_shape[, refcheck]) Change shape and size of array in-place.

round([decimals, out]) Return a with each element rounded to the given number

of decimals.

searchsorted(v[, side, sorter]) Find indices where elements of v should be inserted in

a to maintain order.

setfield(val, dtype[, offset]) Put a value into a speciﬁed place in a ﬁeld deﬁned by a

data-type.

setflags([write, align, uic]) Set array ﬂags WRITEABLE, ALIGNED, (WRITE-

BACKIFCOPY and UPDATEIFCOPY), respectively.

sort([axis, kind, order]) Sort an array, in-place.

squeeze([axis]) Remove single-dimensional entries from the shape of a.

std([axis, dtype, out, ddof, keepdims]) Returns the standard deviation of the array elements

along given axis.

sum([axis, dtype, out, keepdims]) Return the sum of the array elements over the given axis.

swapaxes(axis1, axis2) Return a view of the array with axis1 and axis2 inter-

changed.

take(indices[, axis, out, mode]) Return an array formed from the elements of a at the

given indices.

tobytes([order]) Construct Python bytes containing the raw data bytes in

the array.

tofile(ﬁd[, sep, format]) Write array to a ﬁle as text or binary (default).

tolist() Return the array as a (possibly nested) list.

tostring([order]) Construct Python bytes containing the raw data bytes in

the array.

trace([offset, axis1, axis2, dtype, out]) Return the sum along diagonals of the array.

transpose(*axes) Returns a view of the array with axes transposed.

var([axis, dtype, out, ddof, keepdims]) Returns the variance of the array elements, along given

axis.

Continued on next page

1.1. The N-dimensional array (ndarray) 15

剩余1365页未读，继续阅读

_icrazy_

粉丝: 96
资源: 8

NumPy 1.14.2官方文档：数组操作与通用函数详解

Python库numpy-stl 1.3.4版本发布

numpy-groupies：Matlab中组索引操作的简易代码与优化

深度学习利器：下载numpy-1.26.4版本

Python库 | s2aio-1.14-py3-none-any.whl

PyPI 官网下载 | bottle-neck-1.14.tar.gz

numpy1.14 gcc 版本

tensorflow 1.14 Mac 系统本地跑需要的dylib

pandas官方文档 0.19.2

Numpy.Testing异常测试：处理和测试代码中的异常情况（异常处理指南）

【深度学习加速器】：如何通过Anaconda API文档利用深度学习功能？

最新资源