对于每一行获取最频繁值的频率

Question

I have a dataframe that lookS like this:

var1  var2   var3  var4
a      a      a    b
c      c      b    d
e      e      f    g 
g      a      a    z
g      a      a    g
w      w      w    w

what I want to do is to identify the most frequent value for each row and count the number of times it appears, in this case I'd get

var1  var2   var3  var4  frq
a      a      a    b      3
c      c      b    d      2
e      e      f    g      2
g      a      a    z      2
g      a      a    g      2
w      w      w    w      4

I was thinking to use something like pd.get_dummies but there would be too many dummies as each var1, var2 etc may assume quite a few different values

Answer 1

在转置数据帧上尝试pd.value_counts和max()函数：

df["frq"] = df.T.apply(pd.value_counts).max()

Answer 2

Another way is to apply with axis=1:

df['frq'] = df.apply(lambda x: x.value_counts().iloc[0], axis=1)

Or use stack and groupby :

df['frq'] = df.stack().groupby(level=0).value_counts().max(level=0)

Answer 3

You can use df.mode here. One thing to note is mode finds if there are multiples values.

From pandas-doc df.mode

The mode of a set of values is the value that appears most often. It can be multiple values.

df['frq'] = df.eq(df.mode(axis=1)[0], axis=0).sum(1)

 var1 var2 var3 var4  frq
0    a    a    a    b    3
1    c    c    b    d    2
2    e    e    f    g    2
3    g    a    a    z    2
4    g    a    a    g    2
5    w    w    w    w    4

Answer 4

Here is a sample. I used value_counts and applied it to all rows. I got the count of all variable and then I just choose the maximum value to get the exact result your looking for:

import pandas as pd

df = pd.DataFrame({'var1': ["a","c","e","g","g","w"],
                   'var2': ["a","c","e","a","a","w"],
                   'var3': ["a","b","f","a","a","w"],
                   'var4': ["b","d","g","z","g","w"]})

frequency = df.apply(pd.value_counts, axis=1).max(axis=1)

df["frq"] = frequency

print(df)

Answer 5

Let us try

from scipy import stats
stats.mode(df.values.T)[1]
Out[143]: array([[3, 2, 2, 2, 2, 4]])
#df["frq"] =  stats.mode(df.values.T)[1][0]