In [1]:  #Descriptive Statistics is the building block of data science.

Advanc
​
#Measures of central tendency include mean, median, and the mode, whi
​
#We will cover the topics given below:
​
#Mean
#Median
#Mode
#Standard Deviation
#Variance
#Interquartile Range
#Skewness

In [1]:  import pandas as pd

import numpy as np
import statistics as st
​
# Load the data
df = pd.read_csv("loan_data.csv")
print(df.shape)
print(df.info())

(614, 14)
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 614 entries, 0 to 613
Data columns (total 14 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 Loan_ID 614 non-null object
1 Gender 601 non-null object
2 Age 513 non-null float64
3 Married 611 non-null object
4 Dependents 599 non-null object
5 Education 614 non-null object
6 Self_Employed 582 non-null object
7 ApplicantIncome 614 non-null int64
8 CoapplicantIncome 614 non-null float64
9 LoanAmount 592 non-null float64
10 Loan_Amount_Term 600 non-null float64
11 Credit_History 564 non-null float64
12 Property_Area 614 non-null object
13 Loan_Status 614 non-null object
dtypes: float64(5), int64(1), object(8)
memory usage: 67.3+ KB
None
In [3]:  df.mean()

Out[3]: Age 32.101365

ApplicantIncome 5403.459283
CoapplicantIncome 1621.245798
LoanAmount 146.412162
Loan_Amount_Term 342.000000
Credit_History 0.842199
dtype: float64

In [4]:  print(df.loc[:,'Age'].mean())
print(df.loc[:,'ApplicantIncome'].mean())

32.10136452241716
5403.459283387622

In [5]:  df.mean(axis = 1)[0:10]

Out[5]: 0 1249.000000
1 1316.000000
2 575.333333
3 908.000000
4 1088.666667
5 1713.500000
6 722.500000
7 1017.166667
8 1017.833333
9 4092.833333
dtype: float64

In [6]:  df.median(axis=1)

Out[6]: 0 35.0
1 360.0
2 45.5
3 240.0
4 85.5
...
609 48.0
610 45.0
611 246.5
612 116.0
613 79.5
Length: 614, dtype: float64

In [7]:  df.median()
Out[7]: Age 30.0
ApplicantIncome 3812.5
CoapplicantIncome 1188.5
LoanAmount 128.0
Loan_Amount_Term 360.0
Credit_History 1.0
dtype: float64
In [8]:  #to compute a median of a some column
print(df.loc[:,'Age'].median())
print(df.loc[:,'ApplicantIncome'].median())
​
df.median(axis = 1)[0:10]

30.0
3812.5

Out[8]: 0 35.0
1 360.0
2 45.5
3 240.0
4 85.5
5 313.5
6 227.5
7 259.0
8 264.0
9 354.5
dtype: float64

In [9]:  df.mode()

Out[9]: Loan_ID Gender Age Married Dependents Education Self_Employed ApplicantIncome

0 LP001002 Male 25.0 Yes 0 Graduate No 2500.0

1 LP001003 NaN NaN NaN NaN NaN NaN NaN

2 LP001005 NaN NaN NaN NaN NaN NaN NaN

3 LP001006 NaN NaN NaN NaN NaN NaN NaN

4 LP001008 NaN NaN NaN NaN NaN NaN NaN

... ... ... ... ... ... ... ... ...

609 LP002978 NaN NaN NaN NaN NaN NaN NaN

610 LP002979 NaN NaN NaN NaN NaN NaN NaN

611 LP002983 NaN NaN NaN NaN NaN NaN NaN

612 LP002984 NaN NaN NaN NaN NaN NaN NaN

613 LP002990 NaN NaN NaN NaN NaN NaN NaN

614 rows × 14 columns

Measures the Dispersion

In [10]:  #Measure the Standard deviation
df.std()

Out[10]: Age 7.732178

ApplicantIncome 6109.041673
CoapplicantIncome 2926.248369
LoanAmount 85.587325
Loan_Amount_Term 65.120410
Credit_History 0.364878
dtype: float64

In [12]:  print(df.loc[:,'Age'].std())
print(df.loc[:,'ApplicantIncome'].std())
​
#calculate the standard deviation of the first five rows
df.std(axis = 1)[0:10]

7.732178229043358
6109.041673387174

Out[12]: 0 2575.928085
1 1921.240355
2 1195.703252
3 1219.011567
4 2409.946528
5 2430.485610
6 974.539224
7 1373.763650
8 1569.760799
9 6081.668239
dtype: float64

In [13]:  #easure the Variance

df.var()

Out[13]: Age 5.978658e+01

ApplicantIncome 3.732039e+07
CoapplicantIncome 8.562930e+06
LoanAmount 7.325190e+03
Loan_Amount_Term 4.240668e+03
Credit_History 1.331362e-01
dtype: float64

In [14]:  #Measures the Interquartile Range (IQR)

from scipy.stats import iqr
iqr(df['Age'])

Out[14]: nan
In [3]:  print(df.skew())

Age 0.712146
ApplicantIncome 6.539513
CoapplicantIncome 7.491531
LoanAmount 2.677552
Loan_Amount_Term -2.362414
Credit_History -1.882361
dtype: float64

In [4]:  #The skewness values can be interpreted in the following manner:

​
#Highly skewed distribution: If the skewness value is less than −1 or
​
#Moderately skewed distribution: If the skewness value is between −1
​
#Approximately symmetric distribution: If the skewness value is betwe

In [5]:  df.describe()

Out[5]: Age ApplicantIncome CoapplicantIncome LoanAmount Loan_Amount_Term Credit_

count 513.000000 614.000000 614.000000 592.000000 600.00000 564.

mean 32.101365 5403.459283 1621.245798 146.412162 342.00000 0.

std 7.732178 6109.041673 2926.248369 85.587325 65.12041 0.

min 24.000000 150.000000 0.000000 9.000000 12.00000 0.

25% 25.000000 2877.500000 0.000000 100.000000 360.00000 1.

50% 30.000000 3812.500000 1188.500000 128.000000 360.00000 1.

75% 38.000000 5795.000000 2297.250000 168.000000 360.00000 1.

max 56.000000 81000.000000 41667.000000 700.000000 480.00000 1.

In [6]:  df.describe(include='all')

Out[6]: Loan_ID Gender Age Married Dependents Education Self_Employed Applican

count 614 601 513.000000 611 599 614 582 614

unique 614 2 NaN 2 4 2 2

top LP001157 Male NaN Yes 0 Graduate No

freq 1 489 NaN 398 345 480 500

mean NaN NaN 32.101365 NaN NaN NaN NaN 5403

std NaN NaN 7.732178 NaN NaN NaN NaN 6109

min NaN NaN 24.000000 NaN NaN NaN NaN 150

25% NaN NaN 25.000000 NaN NaN NaN NaN 2877

50% NaN NaN 30.000000 NaN NaN NaN NaN 3812

75% NaN NaN 38.000000 NaN NaN NaN NaN 5795

max NaN NaN 56.000000 NaN NaN NaN NaN 81000

In [7]:  df.groupby('Age').count()

Out[7]: Loan_ID Gender Married Dependents Education Self_Employed ApplicantIncome Coappli

Age

24.0 22 21 21 19 22 20 22

25.0 107 105 107 105 107 102 107

26.0 87 86 87 87 87 81 87

27.0 23 23 22 22 23 22 23

28.0 4 4 4 4 4 2 4

30.0 53 53 53 52 53 52 53

31.0 23 22 23 20 23 23 23

32.0 18 18 18 18 18 18 18

35.0 7 7 7 6 7 7 7

37.0 18 17 18 18 18 16 18

38.0 23 23 23 23 23 23 23

40.0 22 20 22 22 22 20 22

42.0 4 4 4 4 4 4 4

43.0 45 44 45 42 45 40 45

45.0 31 31 30 30 31 30 31

46.0 6 6 6 6 6 5 6

47.0 18 17 18 18 18 18 18

50.0 1 1 1 1 1 1 1

56.0 1 1 1 1 1 1 1

In [8]:  df.groupby('Age')['ApplicantIncome']

Out[8]: <pandas.core.groupby.generic.SeriesGroupBy object at 0x000001CE88961

AF0>
 In [9]:  df.groupby('Age')['ApplicantIncome'].sum()

Out[9]: Age
24.0 109185
25.0 538078
26.0 423267
27.0 178232
28.0 12637
30.0 314638
31.0 174155
32.0 73848
35.0 64332
37.0 99041
38.0 170038
40.0 87414
42.0 28751
43.0 205937
45.0 142359
46.0 38018
47.0 138701
50.0 4106
56.0 6540
Name: ApplicantIncome, dtype: int64

In [12]:  data = {'Gender':['m','f','f','m','f','m','m'],'Age':[24,25,26,27,28,

df_sample = pd.DataFrame(data)
df_sample

Out[12]: Gender Age

0 m 24

1 f 25

2 f 26

3 m 27

4 f 28

5 m 30

6 m 32
In [13]:  f_filter = df_sample['Gender']=='f'
print(df_sample[f_filter])
​
m_filter = df_sample['Gender']=='m'
print(df_sample[m_filter])

Gender Age
1 f 25
2 f 26
4 f 28
Gender Age
0 m 24
3 m 27
5 m 30
6 m 32

In [ ]:  ​