英文:
Build two array from another array based on the values and a window size
问题
我有一个包含一千行和列的数组。数组中有大于1和小于1的数字。我想从中构建两个数组，方式如下：
最重要的部分是小于1的值。然后基于一个窗口大小（这里是7），小于1的值之前大于1的值应该变为1，其余的值都为0。例如，如果一行是[1, 1, 1.2, 0.5, 1.9, 1, 1]，我想要的第一个数组是：[0, 0, 1, 0, 0, 0, 0]，而第二个数组与小于1的值之后大于1的值有关。对于这个例子，我想要 [0, 0, 0, 0, 1, 0, 0]。
这里是一个简单的例子：
我有的数组：
a = np.array([[1,1,1.01, 0.5, 0.5, 1.02, 1, 1,1,1.21, 0.5, 0.5, 1.22, 1.3], [1,1.4,1.01, 0.5, 0.5, 1.02, 1, 1,1,1.51, 0.5, 0.7, 1.22, 1]])
我想要的两个数组：
array([[0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0],
[0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0]])
和
array([[0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 1],
[0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1]])
请问你需要什么帮助？谢谢！
英文:
I have an array with a thousand rows and columns. The array has the number 1, greater than 1, and less than 1. I want to build two arrays from that with this way:
The most important part is the values which are less than 1. Then based on a window size (here is 7), the value greater than 1 before the values less than 1, should change to 1, and all of the other remaining are zero. For example, if a row is [1, 1, 1.2, 0.5, 1.9, 1, 1], the first array that I want is: [0, 0, 1, 0,0,0,0]
and the second array that I want are related to the values greater than 1 after the values less than 1. For this example, I want [0, 0, 0, 0, 1, 0,0].
Here is a simple exmaple:
array I have:
a = np.array([[1,1,1.01, 0.5, 0.5, 1.02, 1, 1,1,1.21, 0.5, 0.5, 1.22, 1.3], [1,1.4,1.01, 0.5, 0.5, 1.02, 1, 1,1,1.51, 0.5, 0.7, 1.22, 1]])
a= array([[1. , 1. , 1.01, 0.5 , 0.5 , 1.02, 1. , 1. , 1. , 1.21, 0.5 ,0.5 , 1.22, 1.3 ],
[1. , 1.4 , 1.01, 0.5 , 0.5 , 1.02, 1. , 1. , 1. , 1.51, 0.5 ,0.7 , 1.22, 1. ]])
Two array I want:
array([[0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0],
[0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0]])
and
array([[0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 1],
[0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1]])
Could you please help me with this? Thank you
答案1
得分: 1
你可以使用scipy中的convolve1d来实现这一点：
from scipy.ndimage import convolve1d
def generate_arrays(a):
# 创建掩模数组
mask1 = np.zeros_like(a)
mask2 = np.zeros_like(a)
# 创建卷积核
kernel = np.ones(7)
# 用卷积核对a进行卷积
convolved = convolve1d(a, kernel, axis=1, mode='constant')
# 创建小于1的值的掩模
mask1[a < 1] = convolved[a < 1] < 1
mask2[a < 1] = 0
# 创建大于1的值的掩模
mask1[a > 1] = 0
mask2[a > 1] = convolved[a > 1] > 1
return mask2, mask1
英文:
You can achieve this by using the convolve1d from scipy:
from scipy.ndimage import convolve1d
def generate_arrays(a):
# create mask arrays
mask1 = np.zeros_like(a)
mask2 = np.zeros_like(a)
# create kernel for convolution
kernel = np.ones(7)
# convolve kernel with a
convolved = convolve1d(a, kernel, axis=1, mode='constant')
# create mask for values less than 1
mask1[a < 1] = convolved[a < 1] < 1
mask2[a < 1] = 0
# create mask for values greater than 1
mask1[a > 1] = 0
mask2[a > 1] = convolved[a > 1] > 1
return mask2, mask1
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论