CODEWITHSUNDEEP
pythonpython-3.xpandasnumpydataframe
BhishanPoudel
BhishanPoudel

Reputation: 17154

Pandas: How to find the binned mean of a column

How can we efficiently find the binned mean of a column in pandas dataframe?

I like to divide the column into 5 parts and find the mean of each part.

Here is what I did:

import numpy as np
import pandas as pd

df = pd.DataFrame({'x': np.arange(20)})
n_bins = 5
dfs = np.array_split(df,n_bins)

x_means = [x.mean()[0] for x in dfs]
n_elems = len(df) // n_bins
x_mean_lst = [[i]*n_elems for i in x_means]
x_mean_array = np.array(x_mean_lst).flatten()
df['x_bin_mean'] = x_mean_array
df

This seems more complicated than necessary. Are there any better alternatives?

The output should look like this:

     x  x_bin_mean
0    0         1.5
1    1         1.5
2    2         1.5
3    3         1.5
4    4         5.5
5    5         5.5
6    6         5.5
7    7         5.5
8    8         9.5
9    9         9.5
10  10         9.5
11  11         9.5
12  12        13.5
13  13        13.5
14  14        13.5
15  15        13.5
16  16        17.5
17  17        17.5
18  18        17.5
19  19        17.5

Upvotes: 2

Views: 109

Answers (1)

cs95
cs95

Reputation: 402533

I'm guessing you want something like

df.groupby(df.index // (len(df) // n_bins))['x'].transform('mean')

or, if your index isn't numeric,

df.groupby(pd.RangeIndex(len(df)) // (len(df) // n_bins))['x'].transform('mean')

Here's what the grouper and output will look like for n_bins = 5,

df.index // (len(df) // 5)
# Int64Index([0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 4, 4], dtype='int64')

df['x_bin_mean'] = (
    df.groupby(df.index // (len(df) // 5))['x'].transform('mean'))
df.head(10)

   x  x_bin_mean
0  0         1.5
1  1         1.5
2  2         1.5
3  3         1.5
4  4         5.5
5  5         5.5
6  6         5.5
7  7         5.5
8  8         9.5
9  9         9.5

Note that integer division, while fast, may not handle cases where the index does not divide equally:

I'm not sure that the integer division is fully correct (if things don't divide evenly). For instance with a length of 16 and n_bins=5 you get 6 groups —Alollz

In this case, use Alollz's helpful suggestion of pd.qcut:

df.groupby(pd.qcut(df.index, n_bins))['x'].transform('mean')

Upvotes: 6

Related Questions