Second, NumPy base: Array modification

One, array properties

• Dimension:. Ndim, returning the current array dimension
• Type:. Dtype, which returns the data type of the elements in the array, note: The array data type defined by NumPy is uniform and cannot be mixed in multiple types
• Shape:. Shape, returns the shape of the data, the number of elements in each layer of the array
• Total number of elements:. Size, returns the total number of elements in the array
• Byte size:. ItemSize, which returns the byte size of each element in the array.

Import NumPy as Npa=np.full ((2,5,4), print ("Array A:", a) print ("Total number of elements in array A:", a.size) print ("Shape of array A:", A.shape) print (" Element type in array a: ", A.dtype) print (" Dimension of Array A: ", A.ndim) print (" element byte size in array A: ", a.itemsize) >>> array A: [[[[]  [[Ten]] [[All] [] [] [] [all]] [ten]] [ten] [all] [] [[All]] [all] [all] ten  10]
   
    [10  [10 10 10 10]] The total number of elements in array A: 40 The shape of the array A: (2, 5, 4) The element type in array A: The dimension of Int32 array A: 3 element in array a byte size: 4
   

Ii. Types of data

• When not specified, NumPy automatically infers the appropriate data type, so it is generally not necessary to display the given data type.
• Numeric Dtype are named by a type name (Eg:int, float, and so on) followed by a number that represents the length of each element
• such as the float data type of python (double-precision floating-point value), takes 8 bytes (64 bits), so it is recorded as float64 in NumPy
• Each data type has a type code, that is, the shorthand method

Data type		Shorthand	Description	Bytes	Range
Integral type	Int_		Default Shaping
	Intc		Equivalent to Long's shaping
	int8	I1	byte shaping	1	[-128,127]
	Int16	I2	Plastic	2	[ -32768,32767]
	Int32	i3	Plastic	4	[ -2^31, 2^31-1]
	Int64	I4	Plastic	6	[ -2^63, 2^63-1]
No sign Integral type	Uint8	U1	No-sign shaping	1	[0,255]
	UInt16	U2	No-sign shaping	2	[0,65535]
	UInt32	U3	No-sign shaping	1	[0, 2^32-1]
	UInt64	U4	No-sign shaping	1	[0,2^64-1]
Boolean	Bool_		Boolean value	1	True or False
Floating point Type	Float_		Float64 Shorthand Form	8
	Float16	F2	Semi-precision floating point type	2	1 Sign bit + 5-bit index + 10-bit decimal part
	Float32	F4 or f	Single-precision floating-point type	4	1 Sign bit + 8-bit index + 23-bit decimal part
	Float64	F8 or d	Double-precision floating-point type	8	1 sign bit + 11-bit index + 52-bit decimal part
Plural	Complex_	C16	shorthand form of complex128
	Complex64	C8	A complex number, represented by two 32-bit floating-point numbers	32
	complex128	C16	A complex number, represented by two 64-bit floating-point numbers	64
Object	Object	O	Python Object Type
String	String_	S	Fixed-length string type (1 bytes per character)
	Unicode_	U	string of fixed-length Unicode type

Third, array modification (attributes)

• Shape modification:. Reshape (),. T
• Dimension modification:. Reshape ()
• Type modification:. Astype ()

3.1, Reshape () method: Modify array shape and dimension

• Directly modifies the shape value of the array ndarray, requiring the product to be unchanged after modification.
• Use the reshape function directly to create a new array that changes size, the shape of the original array remains the same, but the new array and the original array share a memory space , that is, modifying the values in any one array will affect the other. Additionally, the number of elements in the new array is required to match the original array.
• When you specify an axis of 1, the length value of the axis is automatically calculated based on the number of elements in the array.

Import NumPy as Npa=np.arange print ("Original array (one dimension):", a) B=a.reshape (2,3,5) print ("New Array (three-dimensional):", b) B[1][2][2]=666print (" Modifying the values in B affects the A: ", a) print (" Use dimension strike against B: ", B.reshape (10,-1)) >>> original array (one dimension): [0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29] New Array (three-dimensional): [[[0] 1 2  3  4]< c13/>[5  6  7  8  9]  [  [All  ] [[] [[] [] [25 26 27 28 29] ]] Modify the values in B to affect a: [  0   1   2   3   4 5 6   7   8 9-  12 For all the above,  the  666  29] Use a reduced-dimension strike on B: [[  0   1   2] [  3 4 5   ] [  6   7 8   ] [  9 [[] [] [[] [] [] [] []  C71/>26] [666  29]]

3.2. T method: Row and column transpose

• After using the T method, you can transpose the rows and columns of the array.

Import NumPy as Npa=np.random.randint (1,10,size= (2,3,4)) B=a.tprint (B.shape) >>> (4, 3, 2)

3.2. Astype () Method: Modify the element type

• If you need to change the data type of an already existing array, you can modify the Astype method to get a new array.

Import NumPy as Npa=np.arange (1,5) print ("original array: {0} with data type: {1}". Format (A,a.dtype)) B=a.astype (float) print ("new array: {0}, Its data type is: {1} ". Format (B,b.dtype)) >>> original array: [1 2 3 4], whose data type is: Int32 new array: [1. 2.3. 4.], whose data type is: float64

Second, NumPy base: Array modification