In [1]:
# This Python 3 environment comes with many helpful analytics libraries installed
# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python
# For example, here's several helpful packages to load
import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
# Input data files are available in the read-only "../input/" directory
# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory
import os
for dirname, _, filenames in os.walk('/kaggle/input'):
for filename in filenames:
print(os.path.join(dirname, filename))
# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using "Save & Run All"
# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session
Apriori Algorithm Intuition:¶
What are the three essential relations between the support, confidence and lift?
Given two Apps A1 and A2, here are the three essential relations to remember:
Relation between the support and the confidence: confidence(A1 -> A2) =support(A1,A2)/support(A1)
Relation between the lift and the support: lift(A1 -> A2) =support(A1,A2)/support(A1)*support(A2)
Relation between the lift and the confidence (consequence of the two previous equations): lift(A1 -> A2) =confidence(A1 -> A2)/support(A2)
Importing Libraries¶
In [2]:
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
Importing Data¶
In [3]:
dataset = pd.read_csv('/kaggle/input/apps-user-used-the-most/apriori_data.csv')
dataset
Out[3]:
Convert and split the data(String) into the lists format¶
In [4]:
df = list(dataset["AppName"].apply(lambda x:x.split(',')))
dataset=pd.DataFrame(df)
EDA - Exploratory Data Analysis¶
In [5]:
#Displaying the First Five Rows
dataset.head()
Out[5]:
In [6]:
#Displaying the last Five Rows
dataset.tail()
Out[6]:
In [7]:
dataset.shape
Out[7]:
In [8]:
dataset.dtypes
Out[8]:
In [9]:
dataset.isnull().sum()
Out[9]:
In [10]:
dataset.info()
Exporting DataFrame To Csv File¶
In [11]:
dataset.to_csv('apps_dataset.csv',header=None,index=False)
Data Preprocesssing¶
In [12]:
dataset = pd.read_csv('./apps_dataset.csv', header = None)
transactions = []
for i in range(0, 31):
transactions.append([str(dataset.values[i,j]) for j in range(0, 10)])
In [13]:
dataset.head()
Out[13]:
Training the Apriori model on the dataset¶
In [14]:
!pip install apyori
In [15]:
from apyori import apriori
rules = apriori(transactions=transactions, min_support = 0.03, min_confidence = 0.2, min_lift = 2, min_length = 2, max_length = 2)
Visualising the results¶
Displaying the first results coming directly from the output of the apriori function¶
In [16]:
results=list(rules)
In [17]:
results
Out[17]:
Putting the results well organised into a Pandas DataFrame¶
In [18]:
def inspect(results):
App1 = [tuple(result[2][0][0])[0] for result in results]
App2 = [tuple(result[2][0][1])[0] for result in results]
Supports = [result[1] for result in results]
Confidences = [result[2][0][2] for result in results]
Lifts = [result[2][0][3] for result in results]
return list(zip(App1, App2,Supports, Confidences, Lifts))
final_result = pd.DataFrame(inspect(results), columns = ['App 1', 'App 2', 'Support', 'Confidence', 'Lift'])
In [19]:
final_result
Out[19]:
Displaying the results sorted by descending lifts¶
In [20]:
final_result.nlargest(n = 10, columns = 'Lift')
Out[20]:
Conclusion¶
In [21]:
final = final_result.rename({'App 1':'App1','App 2':'App2'}, axis=1)
final
Out[21]:
In [22]:
plt.figure(figsize=(10,10))
sns.heatmap(data=final.corr(), annot=True, cmap='copper_r')
Out[22]:
In [23]:
plt.scatter(final.App1,final.App2)
plt.xlabel('App 1')
plt.ylabel('App 2')
plt.title('App Recommended')
Out[23]:
