knitr::opts_chunk$set(echo = FALSE)
library(readr)
library(ggplot2)
library(dplyr)

## 
## Attaching package: 'dplyr'

## The following objects are masked from 'package:stats':
## 
##     filter, lag

## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union

library(plotly)

## 
## Attaching package: 'plotly'

## The following object is masked from 'package:ggplot2':
## 
##     last_plot

## The following object is masked from 'package:stats':
## 
##     filter

## The following object is masked from 'package:graphics':
## 
##     layout

library(plotrix)

spotify = read.csv("spotify_churn_dataset.csv")

summary(spotify)

##     user_id        gender               age          country         
##  Min.   :   1   Length:8000        Min.   :16.00   Length:8000       
##  1st Qu.:2001   Class :character   1st Qu.:26.00   Class :character  
##  Median :4000   Mode  :character   Median :38.00   Mode  :character  
##  Mean   :4000                      Mean   :37.66                     
##  3rd Qu.:6000                      3rd Qu.:49.00                     
##  Max.   :8000                      Max.   :59.00                     
##  subscription_type  listening_time  songs_played_per_day   skip_rate     
##  Length:8000        Min.   : 10.0   Min.   : 1.00        Min.   :0.0000  
##  Class :character   1st Qu.: 81.0   1st Qu.:25.00        1st Qu.:0.1500  
##  Mode  :character   Median :154.0   Median :50.00        Median :0.3000  
##                     Mean   :154.1   Mean   :50.13        Mean   :0.3001  
##                     3rd Qu.:227.0   3rd Qu.:75.00        3rd Qu.:0.4500  
##                     Max.   :299.0   Max.   :99.00        Max.   :0.6000  
##  device_type        ads_listened_per_week offline_listening   is_churned    
##  Length:8000        Min.   : 0.000        Min.   :0.0000    Min.   :0.0000  
##  Class :character   1st Qu.: 0.000        1st Qu.:0.0000    1st Qu.:0.0000  
##  Mode  :character   Median : 0.000        Median :1.0000    Median :0.0000  
##                     Mean   : 6.944        Mean   :0.7478    Mean   :0.2589  
##                     3rd Qu.: 5.000        3rd Qu.:1.0000    3rd Qu.:1.0000  
##                     Max.   :49.000        Max.   :1.0000    Max.   :1.0000

ggplot(spotify, aes(x=age)) + geom_bar(fill= "lightblue") +
  labs(title = "Age of users of Spotify", x = "Age", y = "Count") + theme_minimal()

xax <- list(title = "Subscription Type", 
    titlefont = list(family = "Verdana"))
yax <- list(title = "Age",               
    titlefont = list(family = "Verdana"))

agesub = plot_ly(
  data  = spotify,
  x = ~subscription_type,
  y = ~age,
  type = "box",
  color = ~subscription_type ) %>%
  layout(xaxis = xax, yaxis = yax, showlegend = TRUE)

xax <- list(title = "Subscription", 
            titlefont = list(family = "Verdana"))
yax <- list(title = "Users",          
            titlefont = list(family = "Verdana"))
df <- as.data.frame(table(spotify$subscription_type, spotify$is_churned))
names(df) <- c("subscription_type", "is_churned", "n")

churn_sub = plot_ly(
    data = df,
    x = ~factor(subscription_type),
    y = ~n,
    color = ~factor(is_churned),
    type = "bar"
  ) %>%
  layout(barmode = "stack", xaxis = xax, yaxis = yax)

listen_sub = ggplot(spotify, aes (x = listening_time, 
      fill = factor( subscription_type,
      levels=c("Free","Premium","Student", "Family")))) + 
  geom_histogram(position= "identity",bins = 30, 
                 color = "white") +
        labs(title = "Listening Time", fill = "Subscription",
             x = "Time", y = "Count") +
        scale_fill_brewer(palette="Pastel1")

age_0 = spotify$age[spotify$is_churned==0]
age_1 = spotify$age[spotify$is_churned==1]
churnage = qqplot(age_0, age_1, xlab = "Retained", ylab = "Churned", main = "Retained vs Churned by Age")
abline(0,1, col = rgb(1,0,1), lwd=2)

There are so many different plots we can use to see this data! Some are more useful than others. Now let’s figure out some problems…

What is the churn rate for spotify? Based on the plots there seems to be more people retained than churned so that’s great for spotify!

Let’s see one final plot just to be sure…