Spotify-Songs-Recommendations-System_with_python

We developed a song recommendation system for the user with the data we received from our Spotify song dataset. Data set and other applications are given in the description. Have a nice day.

import

import pandas as pd
import numpy as np
import seaborn as sns
import plotly.express as px
import plotly.graph_objects as go
import matplotlib.pyplot as plt
import plotly
from sklearn.preprocessing import StandardScaler
from scipy.spatial import distance
import copy
import warnings

warnings.filterwarnings("ignore")
plotly.offline.init_notebook_mode (connected = True)

Having First Look At The Data

data.head()

cols=list(data.columns[11:])
del cols[7]
df=copy.deepcopy(data)
df.drop(columns=cols,inplace=True)
sns.set(style="ticks", context="talk")
plt.style.use("dark_background")
sns.pairplot(df,corner=True,hue='genre')

px.box(data_frame=data,y='duration_ms',color='genre')

x=list(data.corr().columns)
y=list(data.corr().index)
values=np.array(data.corr().values)
fig = go.Figure(data=go.Heatmap(
    z=values,
    x=x,
    y=y,
                   
    
                   hoverongaps = False))
fig.show()

Let's Make A Recommendation System

data=data.dropna(subset=['song_name'])

Creating a new dataframe with required features

# Making a weight matrix using euclidean distance
def make_matrix(data,song,number):
    df=pd.DataFrame()
    data.drop_duplicates(inplace=True)
    songs=data['song_name'].values
#    best = difflib.get_close_matches(song,songs,1)[0]
    best=find_word(song,songs)
    print('The song closest to your search is :',best)
    genre=data[data['song_name']==best]['genre'].values[0]
    df=data[data['genre']==genre]
    x=df[df['song_name']==best].drop(columns=['genre','song_name']).values
    if len(x)>1:
        x=x[1]
    song_names=df['song_name'].values
    df.drop(columns=['genre','song_name'],inplace=True)
    df=df.fillna(df.mean())
    p=[]
    count=0
    for i in df.values:
        p.append([distance.euclidean(x,i),count])
        count+=1
    p.sort()
    for i in range(1,number+1):
        print(song_names[p[i][1]])

a=input('Please enter The name of the song :')
b=int(input('Please enter the number of recommendations you want: '))
make_matrix(df,a,b)

def make_matrix_correlation(data,song,number):
   df=pd.DataFrame()
   data.drop_duplicates(inplace=True)
   songs=data['song_name'].values
#    best = difflib.get_close_matches(song,songs,1)[0]
   best=find_word(song,songs)
   print('The song closest to your search is :',best)
   genre=data[data['song_name']==best]['genre'].values[0]
   df=data[data['genre']==genre]
   x=df[df['song_name']==best].drop(columns=['genre','song_name']).values
   if len(x)>1:
       x=x[1]
   song_names=df['song_name'].values
   df.drop(columns=['genre','song_name'],inplace=True)
   df=df.fillna(df.mean())
   p=[]
   count=0
   for i in df.values:
       p.append([distance.correlation(x,i),count])
       count+=1
   p.sort()
   for i in range(1,number+1):
       print(song_names[p[i][1]])

e=input('Please enter The name of the song :')
f=int(input('Please enter the number of recommendations you want: '))
make_matrix_correlation(df,e,f)

See you on other data science projects. Take care of yourselves:)