如何将矩阵整数转为浮点数？浮点矩阵与象限均值计算问题

我编写了如下Python代码：

import numpy as np

n = float(input("Starting number : " ))
m = float(input("Ending Number : "))

a = np.random.randint(n,m, size=(10,10))
print("\na : ", a)

top_left =  [a[0][:5], a[1][:5]]
top_right = [a[0][5:], a[1][5:]]
bot_left =  [a[2][:5], a[3][:5]]
bot_right = [a[2][5:], a[3][5:]]

print("\n")
print("1st quadrant : ", top_right)
print("2nd quadrant : ", top_left)
print("3rd quadrant : ", bot_left)
print("4th quadrant : ", bot_right)

print("\n")
avg1 = np.average(top_right)
print("Average of 1st quadrant : " ,avg1)

avg2 = np.average(top_left)
print("Average of 2nd quadrant : " ,avg2)

avg3 = np.average(bot_left)
print("Average of 3rd quadrant : " ,avg3)

avg4 = np.average(bot_right)
print("Average of 4th quadrant : " ,avg4)

代码功能为生成10×10矩阵，矩阵数值范围由用户输入的两个数决定，随后划分四个象限并计算各象限均值。我希望支持输入两个浮点数，生成含浮点数的矩阵并以浮点数形式计算均值，但无法实现，请问遗漏了什么？

问题原因 #

你用错了随机数生成函数：np.random.randint() 仅能生成整数类型的随机数，即便输入的n和m是浮点数，它也会自动截断为整数来生成结果，所以矩阵里全是整数，自然无法得到浮点数矩阵。

解决方案 #

1. 替换随机数生成函数：使用np.random.uniform(n, m, size=(10,10))，该函数支持浮点数范围输入，生成的矩阵元素为浮点数。
2. 优化象限切片方式：用numpy数组的切片语法替代手动列表嵌套，代码更简洁且符合numpy的使用规范。

修改后的完整代码：

import numpy as np

n = float(input("Starting number : " ))
m = float(input("Ending Number : "))

# 生成指定范围的浮点数矩阵
a = np.random.uniform(n, m, size=(10,10))
print("\na : ", a)

# 用numpy切片直接获取象限（更简洁规范）
top_left = a[:2, :5]    # 前2行，前5列
top_right = a[:2, 5:]   # 前2行，后5列
bot_left = a[2:4, :5]   # 第3-4行，前5列
bot_right = a[2:4, 5:]  # 第3-4行，后5列

print("\n")
print("1st quadrant : ", top_right)
print("2nd quadrant : ", top_left)
print("3rd quadrant : ", bot_left)
print("4th quadrant : ", bot_right)

print("\n")
avg1 = np.average(top_right)
print("Average of 1st quadrant : ", avg1)

avg2 = np.average(top_left)
print("Average of 2nd quadrant : ", avg2)

avg3 = np.average(bot_left)
print("Average of 3rd quadrant : ", avg3)

avg4 = np.average(bot_right)
print("Average of 4th quadrant : ", avg4)

现在输入浮点数（比如2.5和7.8）就能生成浮点数矩阵，均值也会以浮点数形式输出。

内容的提问来源于stack exchange，提问作者ceilowens