1. Numpy 基礎
1.1. numpy.ndarray 基礎
1.1.1. numpy.ndarray 属性要素
numpy.ndarray 属性要素 | 取得内容 | Example 1 np.array([[1,0,0],[0,1,2]]) |
---|---|---|
ndim | 次元数 | 2 |
shape | 配列の形 | (2,3) |
size | 配列要素の数 | 6 |
dtype | 配列要素のデータ型 | int32 |
T | 転置配列 | np.array([[1 0] [0 1] [0 2]]) |
flags | メモリーレイアウト | |
flat | 一次元化(平坦化)した配列生成 配列定義例:np.array(配列変数.flat) |
|
imag | 配列要素の虚部値配列 | |
real | 配列要素の実部値配列 | |
itemsize | 配列要素の大きさ(バイト) 例:int32 -> 32/8 = 4バイト |
4 |
nbytes | 配列の大きさ(バイト) | 32 |
strides | 隣する配列要素のずれ(バイト) | (12,4) 縦方向:12バイト -> 4バイト * 3要素(横方向) 横方向:4バイト -> 4バイト * 1要素 |
ctypes | ctypesモジュールで使用 | |
base | 参照先の配列 | None |
Example_1.py
### ライブラリ定義
import numpy as np
### 関数定義
def print_attribute(input):
print("")
for key, value in input.items():
print(">>> " + str(key))
print("IN: print("+str(key)+")")
print("OUT: "+str(value))
print("")
### 配列定義
array = np.array( [[1,0,0],[0,1,2]])
### numpy.ndarray要素定義
attribute ={}
attribute['array.ndim'] = array.ndim
attribute['array.shape'] = array.shape
attribute['array.size'] = array.size
attribute['array.dtype'] = array.dtype
attribute['array.T'] = array.T
attribute['array.flags'] = array.flags
attribute['array.flat'] = array.flat
attribute['np.array(array.flat)'] = np.array(array.flat)
attribute['array.imag'] = array.imag
attribute['array.real'] = array.real
attribute['array.itemsize'] = array.itemsize
attribute['array.nbytes'] = array.nbytes
attribute['array.strides'] = array.strides
attribute['array.ctypes'] = array.ctypes
attribute['array.base'] = array.base
### numpy.ndarray要素取得例
print_attribute(attribute)
"""
>>> array.ndim
IN: print(array.ndim)
OUT: 2
>>> array.shape
IN: print(array.shape)
OUT: (2, 3)
>>> array.size
IN: print(array.size)
OUT: 6
>>> array.dtype
IN: print(array.dtype)
OUT: int32
>>> array.T
IN: print(array.T)
OUT: [[1 0]
[0 1]
[0 2]]
>>> array.flags
IN: print(array.flags)
OUT: C_CONTIGUOUS : True
F_CONTIGUOUS : False
OWNDATA : True
WRITEABLE : True
ALIGNED : True
WRITEBACKIFCOPY : False
UPDATEIFCOPY : False
>>> array.flat
IN: print(array.flat)
OUT: <numpy.flatiter object at 0x000001E00DB70A00>
>>> np.array(array.flat)
IN: print(np.array(array.flat))
OUT: [1 0 0 0 1 2]
>>> array.imag
IN: print(array.imag)
OUT: [[0 0 0]
[0 0 0]]
>>> array.real
IN: print(array.real)
OUT: [[1 0 0]
[0 1 2]]
>>> array.itemsize
IN: print(array.itemsize)
OUT: 4
>>> array.nbytes
IN: print(array.nbytes)
OUT: 24
>>> array.strides
IN: print(array.strides)
OUT: (12, 4)
>>> array.ctypes
IN: print(array.ctypes)
OUT: <numpy.core._internal._ctypes object at 0x000001E00D84BC50>
>>> array.base
IN: print(array.base)
OUT: None
"""