Left Join raw and hits

# This scripts will:
# a. create a left join between the raw data (songs) and the hits
# b. filter out all those observations where 'year' is not in 'year' hits table

import pandas as pd 
import numpy as np 
from pandas import ExcelWriter 

#defining the paths to the datasets 
path_raw = r"C:\\Users\\giovanni\\Desktop\\Spotify Final Project\data.xlsx"
path_hits =  r"C:\\Users\\giovanni\\Desktop\\Spotify Final Project\datasets_456949_861300_top10s.xlsx"

#assigning the dataframe to the related variables 
raw = pd.read_excel(path_raw)
hits = pd.read_excel(path_hits)

#Dataset exploration
print("Shape of Spotify data",raw.shape)
print("Shape of Spotify Hits list",hits.shape)
min_date= min(hits['year'])
max_date= max(hits['year'])
count_raw = raw.shape[0]
print("Date range for Hits dataset: \n", min_date, '--',max_date)

##First step###
print("Filtering out from the raw data years which do not fall into the hits year range")
year_list = [] 
for year in hits['year']: 
    year_list.append(year)

raw_filtered = raw[raw['year'].isin(year_list)]
count_raw_filtered=raw_filtered.shape[0]
raw_diff= count_raw -count_raw_filtered
print("Total deleted observations: ", raw_diff)

#I want to join the 2 tables on titles and name
#I will first drop all the columns from hits that are not relevant and keep only the titles
#For every match, we will have a hit
#Note: songs in raw are 'name' whereas in hits they are 'title'
hits=hits['title']

df_merged = pd.merge(raw_filtered, hits, left_on='name', right_on='title',how = 'left')

#adding boolean values to Hits 
#First I need to replace NaN
df_merged['title'].fillna('', inplace = True)
df_merged['Hits'] = np.where(df_merged['title']=='',False,True)

#Writing the new dataset 
with pd.ExcelWriter(r"C:\\Users\\giovanni\\Desktop\\Spotify Final Project\\Mario_Iuliano_FP\raw_data_cleaned.xlsx") as writer:
    df_merged.to_excel(writer,index= False )