Python Numpy recarray.max()函数

在numpy中，数组可以有一个包含字段的数据类型，类似于电子表格中的列。一个例子是[(a, int), (b, float)] ，其中数组中的每个条目是一对(int, float)。通常情况下，这些属性使用字典查询，如arr[‘a’] 和 arr[‘b’] 。记录数组允许字段作为数组的成员被访问，使用arr.a和arr.b。

numpy.recarray.max()函数返回记录数组的最大值或沿轴的最大值。

语法: numpy.recarray.max(axis=None, out=None, keepdims=False)

参数:
axis : [None or int or tuple of ints, optional] 用于操作的一个或多个坐标轴。默认情况下，使用扁平化的输入。
out : [ndarray, optional] 一个储存结果的位置。
-> 如果提供，它必须有一个输入广播到的形状。
-> 如果没有提供或没有，将返回一个新分配的数组。
keepdims : [bool, optional] 如果设置为 “True”，被缩小的轴将作为尺寸为1的尺寸留在结果中。有了这个选项，结果将正确地针对输入阵列进行广播。

返回 : [ndarray or scalar] 记录数组的最大值。如果 axis 是 None，结果是一个标量值。如果给定轴，结果是一个arr.ndim – 1维度的数组。

代码#1：

# Python program explaining
# numpy.recarray.max() method 
  
# importing numpy as geek
import numpy as geek
  
# creating input array with 2 different field 
in_arr = geek.array([[(5.0, 2), (3.0, -4), (6.0, 8)],
                     [(9.0, 1), (5.0, 4), (-12.0, -7)]],
                     dtype =[('a', float), ('b', int)])
  
print ("Input array : ", in_arr)
  
# convert it to a record array,
# using arr.view(np.recarray)
rec_arr = in_arr.view(geek.recarray)
print("Record array of float: ", rec_arr.a)
print("Record array of int: ", rec_arr.b)
  
# applying recarray.max methods
# to float record array along default axis 
# i, e along flattened array
out_arr1 = rec_arr.a.max()
# Maximum of the flattened array 
print("\nMax of float record array, axis = None : ", out_arr1) 
  
  
# applying recarray.max methods
# to float record array along axis 0
# i, e along vertical
out_arr2 = rec_arr.a.max(axis = 0)
# Maximum along 0 axis
print("\nMax of float record array, axis = 0 : ", out_arr2)
  
  
# applying recarray.max methods
# to float record array along axis 1
# i, e along horizontal
out_arr3 = rec_arr.a.max(axis = 1)
# Maximum along 0 axis
print("\nMax of float record array, axis = 1 : ", out_arr3)
  
  
# applying recarray.max methods
# to int record array along default axis 
# i, e along flattened array
out_arr4 = rec_arr.b.max()
# Maximum of the flattened array 
print("\nMax of int record array, axis = None : ", out_arr4) 
  
  
# applying recarray.max methods
# to int record array along axis 0
# i, e along vertical
out_arr5 = rec_arr.b.max(axis = 0)
# Maximum along 0 axis
print("\nMax of int record array, axis = 0 : ", out_arr5)
  
  
# applying recarray.max methods
# to int record array along axis 1
# i, e along horizontal
out_arr6 = rec_arr.b.max(axis = 1)
# Maximum along 0 axis
print("\nMax of int record array, axis = 1 : ", out_arr6)

输出:

Input array :  [[(  5.,  2) (  3., -4) (  6.,  8)]
 [(  9.,  1) (  5.,  4) (-12., -7)]]
Record array of float:  [[  5.   3.   6.]
 [  9.   5. -12.]]
Record array of int:  [[ 2 -4  8]
 [ 1  4 -7]]

Max of float record array, axis = None :  9.0

Max of float record array, axis = 0 :  [9. 5. 6.]

Max of float record array, axis = 1 :  [6. 9.]

Max of int record array, axis = None :  8

Max of int record array, axis = 0 :  [2 4 8]

Max of int record array, axis = 1 :  [8 4]