Python 在索引行和数据帧中的每一行之间查找公共非空列

我想找出索引行和数据帧之间常见的非空列的百分比。例如：

df（指数）=

和df1=

A   B   C   D   E   F
1   3   5  NaN  7  NaN
NaN 2   4  NaN  3   2
7   6  NaN  5   2  NaN
2  NaN  5  NaN  7   3

我在寻找非空公共列的百分比。因此理想的结果是：

    A   B   C   D   E   F   Common%
    1   3   5  NaN  7  NaN    75
    NaN 2   4  NaN  3   2     75
    7   6  NaN  5   2  NaN    50
    2  NaN  5  NaN  7   3     100

---

提前感谢

如果我正确理解了您的问题，应该这样做：

df['Common%'] = (df == index).sum() / index.shape * 100

---

请注意，只有当

索引

是一个

系列

时，这才有效。如果我正确理解了您的问题，这应该可以：

df['Common%'] = (df == index).sum() / index.shape * 100

请注意，只有当

索引

是一个

系列

时，这才有效。您可以执行以下操作：

  # create mask for non-nan values
v = ~np.isnan(df)
w = ~np.isnan(df1)

# count non zeros of intersection
a = np.count_nonzero(np.logical_and(w, v), axis=1)

# count non zeros of v (number of non-nan values in df)
b = np.count_nonzero(v)

# find common percentage
print((a / b) * 100)

输出

[ 75.  75.  50. 100.]

     A    B    C    D  E    F  common
0  1.0  3.0  5.0  NaN  7  NaN    75.0
1  NaN  2.0  4.0  NaN  3  2.0    75.0
2  7.0  6.0  NaN  5.0  2  NaN    50.0
3  2.0  NaN  5.0  NaN  7  3.0   100.0

可以按如下方式添加新列：

df1['common'] = (a / b) * 100
print(df1)

输出

[ 75.  75.  50. 100.]

     A    B    C    D  E    F  common
0  1.0  3.0  5.0  NaN  7  NaN    75.0
1  NaN  2.0  4.0  NaN  3  2.0    75.0
2  7.0  6.0  NaN  5.0  2  NaN    50.0
3  2.0  NaN  5.0  NaN  7  3.0   100.0

你可以做：

  # create mask for non-nan values
v = ~np.isnan(df)
w = ~np.isnan(df1)

# count non zeros of intersection
a = np.count_nonzero(np.logical_and(w, v), axis=1)

# count non zeros of v (number of non-nan values in df)
b = np.count_nonzero(v)

# find common percentage
print((a / b) * 100)

输出

[ 75.  75.  50. 100.]

     A    B    C    D  E    F  common
0  1.0  3.0  5.0  NaN  7  NaN    75.0
1  NaN  2.0  4.0  NaN  3  2.0    75.0
2  7.0  6.0  NaN  5.0  2  NaN    50.0
3  2.0  NaN  5.0  NaN  7  3.0   100.0

可以按如下方式添加新列：

df1['common'] = (a / b) * 100
print(df1)

输出

[ 75.  75.  50. 100.]

     A    B    C    D  E    F  common
0  1.0  3.0  5.0  NaN  7  NaN    75.0
1  NaN  2.0  4.0  NaN  3  2.0    75.0
2  7.0  6.0  NaN  5.0  2  NaN    50.0
3  2.0  NaN  5.0  NaN  7  3.0   100.0

---

我不明白公共值是什么意思？所以对于df1中的第一行，AB C和E是非空的，而在df（索引）中，A C E和F是非空的。因此，两者之间存在共同点。所以{A，C，E}/{A，C，E，F}=3/4，等于75%，我不明白公共值是什么意思？所以对于df1中的第一行，ab C和E是非空的，而在df（索引）中，A C E和F是非空的。因此，两者之间存在共同点。所以{A，C，E}/{A，C，E，F}=3/4，等于75%