Python numpy:将字符串转换为numpy数组 [英] Python numpy: Convert string in to numpy array

问题描述

我有以下已组合在一起的字符串:

I have following String that I have put together:

v1fColor = '2,4,14,5,0,0,0,0,0,0,0,0,0,0,12,4,0,0,0,0,0,0,0,0,0,0,0,0,0,0,15,6,0,0,0,0,1,0,0,0,0,0,0,0,0,0,20,9,0,0,0,2,2,0,0,0,0,0,0,0,0,0,13,6,0,0,0,1,0,0,0,0,0,0,0,0,0,0,10,8,0,0,0,1,2,0,0,0,0,0,0,0,0,0,17,17,0,0,0,3,6,0,0,0,0,0,0,0,0,0,7,5,0,0,0,2,0,0,0,0,0,0,0,0,0,0,4,3,0,0,0,1,1,0,0,0,0,0,0,0,0,0,6,6,0,0,0,2,3'

我将其视为向量:长话短说，它是图像直方图的前景色:

I am treating it as a vector: Long story short its a forecolor of an image histogram:

我具有以下lambda函数来计算两个图像的余弦相似度，因此我尝试将其转换为numpy.array但失败了:

I have the following lambda function to calculate cosine similarity of two images, So I tried to convert this is to numpy.array but I failed:

这是我的lambda函数

Here is my lambda function

import numpy as NP
import numpy.linalg as LA
cx = lambda a, b : round(NP.inner(a, b)/(LA.norm(a)*LA.norm(b)), 3)

因此，我尝试了以下方法将该字符串转换为numpy数组:

So I tried the following to convert this string as a numpy array:

v1fColor = NP.array([float(v1fColor)], dtype=NP.uint8)

但是我最终遇到以下错误:

But I ended up getting following error:

    v1fColor = NP.array([float(v1fColor)], dtype=NP.uint8)
ValueError: invalid literal for float(): 2,4,14,5,0,0,0,0,0,0,0,0,0,0,12,4,0,0,0,0,0,0,0,0,0,0,0,0,0,0,15,6,0,0,0,0,1,0,0,0,0,0,0,0,0,0,20,9,0,0,0,2,2,0,0,0,0,0,0,0,0,0,13,6,0,0,0,1,0,0,0,0,0,0,0,0,0,0,10,8,0,0,0,1,2,0,0,0,0,0,0,0,0,0,17,17,

推荐答案

我正在写此答案，以便将来参考:我不确定在这种情况下什么是正确的解决方案，但我认为@David Robinson最初发布了是一个正确的答案，原因如下:余弦相似度值不能大于1，当我使用NP.array(v1fColor.split(","), dtype=NP.uint8)选项时，对于两个向量之间的余弦相似度，我得到的分段值都大于1.0.

I am writing this answer so if for any future references: I am not sure what is the correct solution in this case but I think What @David Robinson initially publish was the correct answer due to one reason: Cosine Similarity values can not be greater than one and when I use NP.array(v1fColor.split(","), dtype=NP.uint8) option I get strage values which are above 1.0 for cosine similarity between two vectors.

所以我写了一个简单的示例代码来试用:

So I wrote a simple sample code to try out:

import numpy as np
import numpy.linalg as LA

def testFunction():
    value1 = '2,3,0,80,125,15,5,0,0,0,0,0,0,0,0,0,0,0,0,0,2,4,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,4,3,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0'
    value2 = '2,137,0,4,96,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0'
    cx = lambda a, b : round(np.inner(a, b)/(LA.norm(a)*LA.norm(b)), 3)
    #v1fColor = np.array(map(int,value1.split(',')))
    #v2fColor =  np.array(map(int,value2.split(',')))
    v1fColor = np.array( value1.split(','), dtype=np.uint8 )
    v2fColor = np.array( value2.split(','), dtype=np.uint8 )
    print v1fColor
    print v2fColor
    cosineValue = cx(v1fColor, v2fColor)
    print cosineValue

if __name__ == '__main__':
    testFunction()

如果运行此代码，则应获得以下输出:

if you run this code you should get the following output:

不要让un注释两行，并使用David的初始解决方案运行代码:

Not lets un commented two lines that and run the code with the David's Initial Solution:

v1fColor = np.array(map(int,value1.split(',')))
v2fColor =  np.array(map(int,value2.split(','))) 

请牢记上面的余弦相似度值高于1.0，但是当我们使用map函数并使用int强制转换时，我们得到以下值为正确的值:

Keep in mind as you see above Cosine Similarity Value came up above 1.0 but when we use the map function and use do the int casting we get the following value which is the correct value:

幸运的是，我正在绘制最初获取的值，并且某些余弦值超过1.0，然后将这些向量的输出并在python控制台中手动键入，然后通过我的lambda函数发送并获得正确的值回答，所以我很困惑.然后，我编写了测试脚本以查看发生了什么，并很高兴抓住了这个问题.我不是一名Python专家，无法确切地说明两种方法的作用以给出两个不同的答案.但我将其留给@David Robinson或@mgilson.

Luckily I was plotting the values that I was initially getting and some of the cosine values came above 1.0 and I took the outputs of these vectors and manually typed it in python console, and send it via my lambda function and got the correct answer so I was very confuse. Then I wrote the test script to see whats going on and glad I caught this issue. I am not a python expert to exactly tell what is going on in two methods to give two different answers. But I leave that to either @David Robinson or @mgilson.

这篇关于Python numpy:将字符串转换为numpy数组的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！