Python班级关联

Question

我以前试过这个问题。但也许我得保持简单。我有一个从0到1的jaccard值。我有两个类别，每一个从1到7。这些类别值的每一个组合都会产生另一个jaccard值。有什么方法可以找到分类号和jaccard值之间的某种相关性吗？因此，就像类别1的值1总是具有类别2的值2的高jaccard，或类别1的值2总是高jaccard，无论类别2的值是什么？在

import numpy as np
#[category 1, category 2, jaccard]    
array1 = np.array([[1,1,0.1627]
 [1,2,0.2993]
 [1,3,0.1192]
 [1,4,0.201 ]
 [1,5,0.0678]
 [1,6,0.2354]
 [1,7,0.1921]
 [2,1,0.1627]
 [2,2,0.2993]
 [2,3,0.1192]
 [2,4,0.201 ]
 [2,5,0.0678]
 [2,6,0.2354]
 [2,7,0.1921]
 [3,1,0.1627]
 [3,2,0.2993]
 [3,3,0.1192]
 [3,4,0.201 ]
 [3,5,0.0678]
 [3,6,0.2354]])

Answer 1

Pandas是一个很棒的python包，其中包含大量有用的统计/数据科学函数，例如相关性。在

import pandas as pd
import numpy as np

array1 = np.array([[1,1,0.1627],
[1,2,0.2993],
[1,3,0.1192],
[1,4,0.201 ],
[1,5,0.0678],
[1,6,0.2354],
[1,7,0.1921],
[2,1,0.1627],
[2,2,0.2993],
[2,3,0.1192],
[2,4,0.201 ],
[2,5,0.0678],
[2,6,0.2354],
[2,7,0.1921],
[3,1,0.1627],
[3,2,0.2993],
[3,3,0.1192],
[3,4,0.201 ],
[3,5,0.0678],
[3,6,0.2354]])

df = pd.DataFrame(columns=["cat1", "cat2", "jaccard"], data=array1)
df.corr()

^{pr2}$