Python numpy.isnan()

Python numpy.isnan()函数测试元素是否为NaN，并将结果作为布尔数组返回。

语法 :

numpy.isnan(array [, out])

参数 :

array : [array_like]输入数组或对象的元素，我们需要测试是否为无穷大。
out : [ndarray, optional]输出数组与结果放在一起。它的类型被保留，并且它必须有合适的形状来容纳输出。

返回 :

含有结果的布尔数组。对于标量输入，结果是一个新的布尔值，如果输入是正的或负的无穷大，则值为True；否则值为False。
对于数组输入，结果是一个布尔数组，其形状与输入相同，当输入的相应元素为正或负无穷大时，其值为真。其他情况下的值是假的。

代码 1 :

# Python Program illustrating
# numpy.isnan() method
  
import numpy as geek
 
print("Is NaN : ", geek.isnan(1), "\n")
 
print("Is NaN : ", geek.isnan(0), "\n")
 
# not a number
print("Is NaN : ", geek.isnan(geek.nan), "\n")
 
#  infinity
print("Is NaN : ", geek.isnan(geek.inf), "\n")
 
print("Is NaN : ", geek.isnan(geek.NINF), "\n")
 
x = geek.array([-geek.inf, 0., geek.inf])
y = geek.array([2, 2, 2])
print("Checking for NaN : ", geek.isnan(x, y))

输出 :

Is NaN :  False 

Is NaN :  False 

Is NaN :  True 

Is NaN :  False 

Is NaN :  False 

Checking for NaN :  [0 0 0]

代码 2 :

# Python Program illustrating
# numpy.isnan() method
   
import numpy as geek
  
# Returns True/False value for each element
b = geek.arange(20).reshape(5, 4)
                
print("\n",b)
print("\nIs NaN(Not a Number): \n", geek.isnan(b))
 
 
b = [[1j],
     [geek.nan]]
print("\nIs NaN(Not a Number) : \n", geek.isnan(b))

输出 :

 [[ 0  1  2  3]
 [ 4  5  6  7]
 [ 8  9 10 11]
 [12 13 14 15]
 [16 17 18 19]]

Is NaN(Not a Number): 
 [[False False False False]
 [False False False False]
 [False False False False]
 [False False False False]
 [False False False False]]

Is NaN(Not a Number) : 
 [[False]
 [ True]]