Incorrect value of Agg Mean

Question

I need to get the same result of the media, fashion and average of this table

Televisores/dia   Freq. absoluta
 0 |----- 20           5
 20|----- 40           25
 40|----- 60           40
 60|----- 80           15
 80|----- 100          10
100|----- 120          5

media=53 moda=50 mediana=50

The idea is to calculate the average of each value in the first column and then the frequency of each one. I arrived at this result:

televisores = [*range(0, 120)]
frequencia = [5, 25, 40, 15, 10, 5]

df = pd.DataFrame({'televisores': televisores})

bins = pd.cut(df['televisores'], [0, 20, 40, 60, 80, 100, 120])
df = df.groupby(bins)['televisores'].agg(Media='mean')

df['Freq. absoluta'] = frequencia

count = [x for x,y in zip(df['Media'], df['Freq. absoluta']) for i in range(y)]

The problem is that the media returns the values with 0.5 more

    df
                 Media   Freq. absoluta
    televisores
    (0, 20]       10.5         5
    (20, 40]      30.5         25
    (40, 60]      50.5         40
    (60, 80]      70.5         15
    (80, 100]     90.5         10
    (100, 120]   110.0         5

    mean(count), mode(count), median(count)
      53.475       50.5         50.5

I wanted to understand the problem and know if there is any easier way to get the result.

Answer 1

The script is doing what it should. Assuming the averages are wrong at 0.5 is wrong.

Some points:

1. range(0,120) will create a list starting at zero and ending at 119.
2. pd.cut will create 6 ranges where by default it has the format (x...y]. Where the x side is opened and the y side closed. To flip use: right=False in the cut
3. pd.cut does not include the lower value. to make it use: include_lowest=True

Using the cut, you’d have intervals like

>>> import numpy as np
>>> for sdf in np.array_split(df, 6):
...     print(list(sdf.televisores))
...
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]
[20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39]
[40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59]
[60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79]
[80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99]
[100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119]

Realize that

>>> mean([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19])
9.5

>>> mean([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20])
10

>>> mean([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21])
10.5

However, in case you use .loc:

>>> import pandas as pd
>>> import numpy as np
>>> from statistics import mean, median, mode

>>> televisores = [*range(0, 120)]
>>> frequencia = [5, 25, 40, 15, 10, 5]

>>> df = pd.DataFrame({'televisores': televisores})

>>> df.loc[0:20].mean()
televisores    10.0
dtype: float64
>>> df.loc[20:40].mean()
televisores    30.0
dtype: float64
>>> df.loc[40:60].mean()
televisores    50.0
dtype: float64
>>> df.loc[60:80].mean()
televisores    70.0
dtype: float64
>>> df.loc[80:100].mean()
televisores    90.0
dtype: float64

Although you get the averages you were hoping for, this is wrong. In the above case, df.loc[0:20].mean() and df.loc[20:40].mean() use the same 20 for both average calculations.

Note the .loc includes the initial and final indices in the result

>>> df.loc[0:3]
   televisores
0            0
1            1
2            2
3            3