Welcome to 30 Days ML | Day 21

Import Library
In [2]: import numpy as np
import pandas as pd
from sklearn.model_selection import cross_val_score
from sklearn.linear_model import LogisticRegression
import seaborn as sns

Import Dataset
In [3]: df = pd.read_csv('train.csv')

In [4]: df.head()

Out[4]:
PassengerId Survived Pclass Name Sex Age SibSp Parch Ticket Fare Cabin Emb

Braund,
0 1 0 3 Mr. Owen male 22.0 1 0 A/5 21171 7.2500 NaN
Harris

Cumings,
Mrs. John
Bradley
1 2 1 1 female 38.0 1 0 PC 17599 71.2833 C85
(Florence
Briggs
Th...

Heikkinen,
STON/O2.
2 3 1 3 Miss. female 26.0 0 0 7.9250 NaN
3101282
Laina

Futrelle,
Mrs.
Jacques
3 4 1 1 female 35.0 1 0 113803 53.1000 C123
Heath
(Lily May
Peel)

Allen, Mr.
4 5 0 3 William male 35.0 0 0 373450 8.0500 NaN
Henry

In [5]: df = pd.read_csv('train.csv')[['Age','Pclass','SibSp','Parch','Survived']]

In [6]: df.head()

Out[6]:
Age Pclass SibSp Parch Survived

0 22.0 3 1 0 0

1 38.0 1 1 0 1

2 26.0 3 0 0 1

3 35.0 1 1 0 1

4 35.0 3 0 0 0
 Drop NA Value
In [7]: df.dropna(inplace=True)

In [8]: df.sample(5)

Out[8]:
Age Pclass SibSp Parch Survived

663 36.0 3 0 0 0

498 25.0 1 1 2 0

342 28.0 2 0 0 0

136 19.0 1 0 2 1

884 25.0 3 0 0 0

Separate X and Y
In [9]: X = df.iloc[:,0:4]
y = df.iloc[:,-1]

In [10]: X.head()

Out[10]:
Age Pclass SibSp Parch

0 22.0 3 1 0

1 38.0 1 1 0

2 26.0 3 0 0

3 35.0 1 1 0

4 35.0 3 0 0

Check Accuracy for Logistic Regression

In [11]: np.mean(cross_val_score(LogisticRegression(),X,y,scoring='accuracy',cv=20))

Out[11]: 0.6933333333333332

Applying Feature Construction

Create New Column

In [12]: X['Family_size'] = X['SibSp'] + X['Parch'] + 1
In [13]: X.head()

Out[13]:
Age Pclass SibSp Parch Family_size

0 22.0 3 1 0 2

1 38.0 1 1 0 2

2 26.0 3 0 0 1

3 35.0 1 1 0 2

4 35.0 3 0 0 1

Apply New Function

In [14]: def myfunc(num):
if num == 1:
#alone
return 0
elif num >1 and num <=4:
# small family
return 1
else:
# large family
return 2

In [15]: myfunc(4)

Out[15]: 1

Apply M Function
In [16]: X['Family_type'] = X['Family_size'].apply(myfunc)

In [17]: X.head()

Out[17]:
Age Pclass SibSp Parch Family_size Family_type

0 22.0 3 1 0 2 1

1 38.0 1 1 0 2 1

2 26.0 3 0 0 1 0

3 35.0 1 1 0 2 1

4 35.0 3 0 0 1 0

Drop unwanted columns

In [18]: X.drop(columns=['SibSp','Parch','Family_size'],inplace=True)
In [19]: X.head()

Out[19]:
Age Pclass Family_type

0 22.0 3 1

1 38.0 1 1

2 26.0 3 0

3 35.0 1 1

4 35.0 3 0

Review Accuracy after Feature Construction

In [20]: np.mean(cross_val_score(LogisticRegression(),X,y,scoring='accuracy',cv=20))

Out[20]: 0.7003174603174602

Feature Splitting (New Topic)

Review Import Data

In [21]: df = pd.read_csv('train.csv')

In [22]: df.head()

Out[22]:
PassengerId Survived Pclass Name Sex Age SibSp Parch Ticket Fare Cabin Emb

Braund,
0 1 0 3 Mr. Owen male 22.0 1 0 A/5 21171 7.2500 NaN
Harris

Cumings,
Mrs. John
Bradley
1 2 1 1 female 38.0 1 0 PC 17599 71.2833 C85
(Florence
Briggs
Th...

Heikkinen,
STON/O2.
2 3 1 3 Miss. female 26.0 0 0 7.9250 NaN
3101282
Laina

Futrelle,
Mrs.
Jacques
3 4 1 1 female 35.0 1 0 113803 53.1000 C123
Heath
(Lily May
Peel)

Allen, Mr.
4 5 0 3 William male 35.0 0 0 373450 8.0500 NaN
Henry

In [23]: #Use Name Column

In [24]: df['Name']

Out[24]: 0 Braund, Mr. Owen Harris

1 Cumings, Mrs. John Bradley (Florence Briggs Th...
2 Heikkinen, Miss. Laina
3 Futrelle, Mrs. Jacques Heath (Lily May Peel)
4 Allen, Mr. William Henry
...
886 Montvila, Rev. Juozas
887 Graham, Miss. Margaret Edith
888 Johnston, Miss. Catherine Helen "Carrie"
889 Behr, Mr. Karl Howell
890 Dooley, Mr. Patrick
Name: Name, Length: 891, dtype: object

Separate Salutation
In [25]: df['Title'] = df['Name'].str.split(', ', expand=True)[1].str.split('.', expand=True)[0
​

In [26]: ​
df['Name'].str.split(', ', expand=True)[1].str.split('.', expand=True)[0]
​

Out[26]: 0 Mr
1 Mrs
2 Miss
3 Mrs
4 Mr
...
886 Rev
887 Miss
888 Miss
889 Mr
890 Mr
Name: 0, Length: 891, dtype: object

In [27]: df[['Title','Name']]

Out[27]:
Title Name

0 Mr Braund, Mr. Owen Harris

1 Mrs Cumings, Mrs. John Bradley (Florence Briggs Th...

2 Miss Heikkinen, Miss. Laina

3 Mrs Futrelle, Mrs. Jacques Heath (Lily May Peel)

4 Mr Allen, Mr. William Henry

... ... ...

886 Rev Montvila, Rev. Juozas

887 Miss Graham, Miss. Margaret Edith

888 Miss Johnston, Miss. Catherine Helen "Carrie"

889 Mr Behr, Mr. Karl Howell

890 Mr Dooley, Mr. Patrick

891 rows × 2 columns

 Review Analysis after Feature Splitting
In [28]: np.mean(cross_val_score(LogisticRegression(),X,y,scoring='accuracy',cv=20))

Out[28]: 0.7003174603174602