braycurtis不相似矩陣（numpy數組）結果中的一些元素包含'nan'

Question

我計算了以下矩陣的braycurtis不相似矩陣。行是社區ANS列是種

[[ 0 0 0 0] 
[ 13 110 0 0] 
[ 6 3 0 0] 
[ 0 5 0 0] 
[ 0 128 0 0] 
[ 0 0 0 0] 
[ 11 76 11 0] 
[ 8 29 3 0] 
[ 0 58 5 0] 
[ 1 3 0 0] 
[ 4 11 1 0] 
[ 3 38 0 0] 
[ 9 35 8 7] 
[ 0 0 0 0] 
[ 0 0 0 0] 
[ 63 576 11 9] 
[ 24 99 0 0] 
[ 1 29 5 0] 
[ 0 0 0 0] 
[ 0 0 0 0] 
[ 0 0 0 0] 
[ 0 0 0 0] 
[ 2 0 0 0] 
[ 1 4 0 0] 
[ 0 0 0 0] 
[ 8 20 0 0] 
[ 3 21 13 6] 
[ 1 4 0 0] 
[ 0 0 0 0] 
[ 0 0 0 0] 
[ 0 4 0 0]] 

我用

dissimilarityMatrix=scipy.spatial.distance.pdist(CommGrouped, 'braycurtis') 

，並得到以下結果

[ 1.   1.   1.   1.     nan 1.   1. 
    1.   1.   1.   1.   1.     nan 
     nan 1.   1.   1.     nan   nan 
     nan   nan 1.   1.     nan 1.   1. 
    1.     nan   nan 1.   0.86363636 0.921875 
    0.12350598 1.   0.21266968 0.54601227 0.37634409 0.93700787 
    0.78417266 0.5   0.51648352 1.   1.   0.68542199 
    0.08943089 0.62025316 1.   1.   1.   1.   0.968 
    0.921875 1.   0.62913907 0.71084337 0.921875 1.   1. 
    0.93700787 0.57142857 0.95620438 1.   0.8317757 0.63265306 
    0.91666667 0.38461538 0.44  0.76  0.73529412 1.   1. 
    0.97305389 0.86363636 0.81818182 1.   1.   1.   1. 
    0.63636364 0.42857143 1.   0.51351351 0.76923077 0.42857143 
    1.   1.   0.53846154 0.92481203 1.   0.90291262 
    0.77777778 0.85294118 0.33333333 0.52380952 0.7826087 0.84375  1. 
    1.   0.98493976 0.921875 0.75  1.   1.   1. 
    1.   1.   0.2   1.   0.6969697 0.79166667 
    0.2   1.   1.   0.11111111 1.   0.32743363 
    0.6547619 0.39267016 0.95454545 0.84722222 0.55029586 0.62566845 
    1.   1.   0.6747141 0.21115538 0.64417178 1.   1. 
    1.   1.   1.   0.93984962 1.   0.74358974 
    0.75438596 0.93984962 1.   1.   0.93939394 1.   1. 
    1.   1.   1.   1.   1.     nan 
     nan 1.   1.   1.     nan   nan 
     nan   nan 1.   1.     nan 1.   1. 
    1.     nan   nan 1.   0.42028986 0.2173913 
    0.92156863 0.71929825 0.41007194 0.33757962 1.   1. 
    0.74108322 0.21266968 0.47368421 1.   1.   1.   1. 
    0.96  0.90291262 1.   0.55555556 0.5035461 0.90291262 
    1.   1.   0.92156863 0.37864078 0.81818182 0.42857143 
    0.20987654 0.19191919 1.   1.   0.88555079 0.54601227 
    0.12  1.   1.   1.   1.   0.9047619 
    0.77777778 1.   0.17647059 0.34939759 0.77777778 1.   1. 
    0.81818182 0.91044776 0.69620253 0.26923077 0.3442623 1.   1. 
    0.82548476 0.37634409 0.30612245 1.   1.   1.   1. 
    1.   0.88235294 1.   0.56043956 0.50943396 0.88235294 
    1.   1.   0.88059701 0.6   0.82222222 0.87301587 
    1.   1.   0.98793363 0.93700787 0.79487179 1.   1. 
    1.   1.   0.66666667 0.11111111 1.   0.75 
    0.82978723 0.11111111 1.   1.   0.25  0.50877193 
    0.57333333 1.   1.   0.95259259 0.78417266 0.49019608 
    1.   1.   1.   1.   0.77777778 0.52380952 
    1.   0.31818182 0.49152542 0.52380952 1.   1.   0.6 
    0.24  1.   1.   0.88285714 0.5   0.21052632 
    1.   1.   1.   1.   0.90697674 0.7826087 1. 
    0.33333333 0.42857143 0.7826087 1.   1.   0.82222222 
    1.   1.   0.8356546 0.51648352 0.25531915 1.   1. 
    1.   1.   0.93442623 0.84375  1.   0.35632184 
    0.25490196 0.84375  1.   1.   0.87301587   nan 
    1.   1.   1.     nan   nan   nan 
     nan 1.   1.     nan 1.   1.   1. 
     nan   nan 1.   1.   1.   1. 
     nan   nan   nan   nan 1.   1. 
     nan 1.   1.   1.     nan   nan 
    1.   0.68542199 0.89913545 1.   1.   1.   1. 
    0.99394856 0.98493976 1.   0.91848617 0.88319088 0.98493976 
    1.   1.   0.98793363 0.62025316 1.   1.   1. 
    1.   0.968  0.921875 1.   0.62913907 0.71084337 
    0.921875 1.   1.   0.93700787 1.   1.   1. 
    1.   0.94594595 0.75  1.   0.33333333 0.30769231 
    0.75  1.   1.   0.79487179   nan   nan 
     nan 1.   1.     nan 1.   1.   1. 
     nan   nan 1.     nan   nan 1.   1. 
     nan 1.   1.   1.     nan   nan 
    1.     nan 1.   1.     nan 1.   1. 
    1.     nan   nan 1.   1.   1. 
     nan 1.   1.   1.     nan   nan 
    1.   0.71428571 1.   0.86666667 0.91111111 0.71428571 
    1.   1.   1.   1.   0.6969697 0.79166667 
    0.   1.   1.   0.11111111 1.   1.   1. 
     nan   nan 1.   0.35211268 0.6969697 1.   1. 
    0.75  0.79166667 1.   1.   0.82978723 1.   1. 
    0.11111111   nan 1.   1.  ] 

我無法弄清楚，如果結果是正確的和獲得nan的原因。

請幫忙！

Answer 1

是的，輸出是正確的。

這是很容易驗證自己：所述documentation on pdist具有用於佈雷-柯蒂斯距離實際的公式：

d（U，V）=Σ_我（U _我 - v _我）/Σ_我（U _我 + v _我）

所以你可以自己計算所有的距離（或只是幾個檢查）。

NaNs也很明顯：它們是分子和分母均爲0的結果，這是發生在輸入中的一組向量組合中的結果。