ALL_IMPORTANT_LIBRARIES

Statistics and Machine Learning (Simple Linear Regression)

— — — — — — — — — — — — — — — — — — — — — — — —

Last Checkpoint: 10 minutes ago(autosaved)

Python 3

Trusted

• File
• Edit
• View
• Insert
• Cell
• Kernel
• Widgets
• Help

Run

CodeMarkdownRaw NBConvertHeading-

ASAD ASHRAF KARL

PYTHON LIBRARIES

To avoid warnings:

In [ ]:

1

from warnings import filterwarnings

2

filterwarnings('ignore')

Python Fundamental libraries:

In [1]:

1

import numpy as np

2

​

3

import pandas as pd

4

​

5

import math

6

​

7

import random

Visualizations:

In [10]:

1

import matplotlib.pyplot as plt

2

​

3

import seaborn as sns

STATISTICS:

In [3]:

1

import scipy.stats as st

2

​

3

from itertools import combinations

4

# list(combinations(data,x)) , x=1,2,3,.... number of combinations

5

​

6

import statsmodels.api as sm

7

from statsmodels.formula.api import ols

8

# model = ols('numerical_variale'~'Categorical_variable', data).fit()

9

# sm.stats.anova_lm(model)

10

​

11

# If we reject null hypothesis, and wish to all possibilites in difference of mean, we work on another library:

12

from statsmodels.stats.multicomp import pairwise_tukeyhsd

13

# pairwise_tukeyhsd('numerical_variable', 'categorecal_variable', alpha).summery()

14

​

15

# Varias functions from scipy:

16

from scipy import stats

17

from scipy.stats import shapiro

Split

In [ ]:

1

from sklearn.model_selection import train_test_split

2

​

3

x=df.drop('target_column', axis = 1)#drop the target variable

4

y=df['target_column']

5

xtrain, xtest, ytrain, ytest = train_test_split(x, y, test_size = 0.2, random_state = 42)

FEATURE ENGINEERING:

STANDARDIZATION:

z_score

In [12]:

1

from scipy.stats import zscore

2

# df['New_column'] = zscore(df['On_which_column'])

NORMALIZATION:

MinMax scaling

In [5]:

1

from sklearn.preprocessing import MinMaxScaler

2

# mm = MinMaxScaler(feature_range=(0, 1)) To fix theh range according to user's wish.

3

# df['New_column'] = mm.fit_transform(df[['On_which_column']])

OR

In [ ]:

1

df['New_Column'], lmda = st.boxcox(df['Column_name'])

LABEL ENCODING:

In [7]:

1

from sklearn.preprocessing import LabelEncoder

2

# LE = LabelEncoder()

3

# df['New_column'] = LE.fit_transform(df['On_which_column'])

ONE HOT ENCODING:

In [9]:

1

# pd.get_dummies(df['Column_name'])

Various functions from statsmodel to perform linear regression

In [ ]:

1

import statsmodels

2

import statsmodels.api as sm

3

import statsmodels.stats.api as sms

4

from statsmodels.compat import lzip

5

from statsmodels.stats.outliers_influence import variance_inflation_factor

6

from statsmodels.graphics.gofplots import qqplot

7

from statsmodels.stats.anova import anova_lm

8

from statsmodels.formula.api import ols

9

from statsmodels.tools.eval_measures import rmse

10

from statsmodels.formula.api import ols

Sklearn libraries:

In [ ]:

1

from sklearn.model_selection import train_test_split, GridSearchCV

2

from sklearn.linear_model import Linear_regression, Lasso, Ridge, LassoCV, RidgeCV, ElasticNet

3

from sklearn.feature_selection import RFE

4

from sklearn.ensemble import RandomForestRegressor,  BaggingRegressor

5

# RandomForestRegressor Learns the maximum of the data. (Suitable model) , BaggingRegressor for good prediction.

‘metrics’ from sklearn is used for evaluating the model performance

In [ ]:

1

from sklearn.metrics import mean_absolute_error, mean_squared_error, r2_score

END