Numpy - Construction of array

1.Numpy summary

      Standard installation Python Use the list (list) Save a set of values , Can be used as an array , However, since the elements of the list can be any object , So what's in the list is a pointer to the object . So in order to save a simple [1,2,3], Need to have 3 Three pointers and three integer objects . For numerical operation, this structure is obviously a waste of memory and CPU computing time .NumPy The birth of the new century has made up for these shortcomings ,NumPy Two basic objects are provided ：ndarray（N-dimensional array object） and ufunc（universal function object）.ndarray( We call it array ) It's a multidimensional array that stores a single data type , and ufunc Is a function that can process arrays , That is, the built-in function that can perform vectorization .

2. Array array Construction

（1）tuple and list Turn into numpy Of array data type

import numpy as np
# By list list Generating arrays array
my_list=[2,3,5,7,8]
print(my_list)
my_np_array=np.array(my_list)
print(my_np_array)
# By tuples tuple Generating arrays array
my_tuple=(9,6,3,12)
my_np_array1=np.array(my_tuple)
# list 、 The multiple of tuple is different from that of array 
print(my_list*3)
print(my_tuple*3)
print(np.array(my_tuple)*3)

result ：

[2, 3, 5, 7, 8]
[2 3 5 7 8]
[2, 3, 5, 7, 8, 2, 3, 5, 7, 8, 2, 3, 5, 7, 8]
(9, 6, 3, 12, 9, 6, 3, 12, 9, 6, 3, 12)
[27 18  9 36]

（2）arange()、linspace() Generating arrays

import numpy as np
#1. function arange()
np.arange(9)   # This function generates an array array, Excluding mantissa 
np.arange(2,9) # Left closed right away 
print(np.arange(2,9,2) )# [2 4 6 8]
#2. function linspace()
print(np.linspace(2,10,10)) # Within the specified time , Return fixed interval data 

（3） Element is 1 Matrix ones And elements are all 0 Matrix zeros Automatic generation of

np.zeros(6)   # One dimensional zero array 
np.zeros((2,3))  # Two dimensional zero array 
np.ones(5)    # One dimensional array , The elements are 1
np.ones((5,3))   # Two dimensional array , The elements are 1
np.ones(5,dtype='int32')   # Specifies the data type of the array element , The default is floating point .

（4） Random array generation

. Generate random numbers

import numpy as np
np.random.seed(10)  # random.seed(), Omitting parameters means taking the current system time as the seed 
np.random.random()  ##  Produce a [0,1) Floating point random number between 
print(np.random.random())  # The same seed value , The resulting random numbers are the same 

. Generate a random array with normal distribution

#1. Generate standard normal distribution random array 
import numpy as np
np.random.normal(size=(5, 2))  #  produce 5x2 Each element of comes from the sample matrix of standard normal distribution ,
#2. The mean value of the generation is 85, The standard deviation is 4 Of 100 A normally distributed random number 
sampleNo = 100;   # Number of samples 
mu = 85     # mean value 
sigma = 4   # Standard deviation 
s1 = np.random.normal(mu, sigma, size=sampleNo)  # The return is a one-dimensional array 
s2 = np.random.normal(mu, sigma, size=(5,2))     # The return is a two-dimensional array 

3 Example

Construct a 3 That's ok 4 Two dimensional array of columns , The elements in the array come from the average value 2, The standard deviation is 3 Is a normal distribution .

import numpy as np
np.random.seed(20)
mu=2
sigma=3
my_array=np.random.normal(mu, sigma, size=(3,4))
print(my_array)

result ：

[[ 4.65167934  2.58759507  3.07260955 -5.02978572]
 [-1.25449776  3.67908887  4.81840805 -0.93544313]
 [ 3.50929052  3.21924341  2.97038303  0.51976735]]