Netflix Viewing History in R

Overview

This vignette will provide information on Netflix viewers history in R.

Tool

Collected data from Netflix. https://www.netflix.com/YourAccount .

Data source : To download the file, please click here:

[Netflix] (https://drive.google.com/file/d/1jVHjDdsIPaiwpXj6QIt1dATu3Dn1ZTth/view?usp=sharing))

or

https://drive.google.com/file/d/1jVHjDdsIPaiwpXj6QIt1dATu3Dn1ZTth/view?usp=sharing

Vignette Info

While working with the data, it demostrates different use of R functions.

The dataset that I choosed is Netflix Viewing History dataset. This dataset contains 5673 rows.

library(dplyr)
library(tidyr)
library(lubridate)
library(zoo)
library(ggplot2)
library(ggplotlyExtra)
library(plotly)

# READING DATA FROM CSV DOWNLOADED FROM NETFLIX ACCOUNT
netflix <- readr::read_csv("netflix_combined.csv") 
str(netflix)
head(netflix)
netflix$Date <- dmy(netflix$Date)
  

# SEPARATE TITLE COLUMN IN TITLE OF TV SERIES, SEASON AND EPISODE TITLE
netflix_series <- netflix %>%
  separate(col = Title, into = c("title", "types", "title_episode"), sep = ': ')


# REMOVE OCCURRENCES WHERE SEASON AND EPISODE ARE EMPTY (BECAUSE THEY ARE NOT TV SERIES)
netflix_series <- netflix_series[!is.na(netflix_series$types),]
netflix_series <- netflix_series[!is.na(netflix_series$title_episode),]
maratones_netflix <- netflix_series %>%
  count(title, Date

# EPISODES PER DAY
netflix_episode_perday <- netflix %>%
  count(Date) %>%
  arrange(desc(n))

Plotting Eposodes Per Day

netflix_episode_perday_plot <- ggplot(aes(x = Date, y = n, color = n), data = netflix_episode_perday) +
  geom_col(color = c("blue")) +
  theme_minimal() +
  ggtitle("Episodes watched on my Netflix per day", "History from 2020 to 2022") +
  labs(x = "Date", y = "watched episodes") 
netflix_episode_perday_plot

Graph 1: Episodes watched on Netflix per Day

Graph 1 shows that the number of episodes watched annually increased gradually over time. 

Format into Day, month and year

netflix_episode_perday <- netflix_episode_perday[order(netflix_episode_perday$Date),]
netflix_episode_perday$diasemana <- wday(netflix_episode_perday$Date)
netflix_episode_perday$diasemanaF <- weekdays(netflix_episode_perday$Date, abbreviate = T)
netflix_episode_perday$mesF <- months(netflix_episode_perday$Date, abbreviate = T)

Days of the week and month

netflix_episode_perday$diasemanaF <-factor(netflix_episode_perday$diasemana, levels = rev(1:7), labels = rev(c("Mon","Tue","Wed","Thu","Fri","Sat","Sun")),ordered = TRUE)
netflix_episode_perday$mesF <- factor(month(netflix_episode_perday$Date),levels = as.character(1:12), labels = c("Jan","Feb","Mar","Apr","May","Jun","Jul","Aug","Sep","Oct","Nov","Dec"),ordered = TRUE)
netflix_episode_perday$añomes <- factor(as.yearmon(netflix_episode_perday$Date)) 
netflix_episode_perday$semana <- as.numeric(format(netflix_episode_perday$Date,"%W"))
netflix_episode_perday$semanames <- ceiling(day(netflix_episode_perday$Date) / 7)
netflix_episode_perday_calendario <- ggplot(netflix_episode_perday, aes(semanames, diasemanaF, fill = n)) + 
  geom_tile(colour = "white") + 
  facet_grid(year(netflix_episode_perday$Date) ~ mesF) + 
  scale_fill_gradient(low = "blue", high = "green") + 
  ggtitle("Episodes watched per day on Netflix", "Heatmap by day of the week, month and year") +
  labs(x = "week number", y = "Weekday") +
  labs(fill = "No.Episodes")
netflix_episode_perday_calendario

Graph 2: Episodes watched per day on Netflix based on week, month, and year

Graph 2 shows the highest number of episodes watched in Mar 2020 compared to other months.There was a dip in March due to low count views of days collected for that period.

Frequency of activity in my netflix account per day

view_day <- netflix_episode_perday %>%
  count(diasemanaF)
view_dayview_day_plot <- view_day %>% 
  ggplot(aes(diasemanaF, n)) +
  geom_col(fill = "orange") +
  coord_polar()  +
  theme_minimal() +
  theme(axis.title.x = element_blank(),
        axis.title.y = element_blank(),
        axis.text.y = element_blank(),
        axis.text.x = element_text(face = "bold"),
        plot.title = element_text(size = 16, face = "bold")) +
  ggtitle("Frequency of episodes watched", "Activity by day of the week on Netflix")

view_dayview_day_plot

Graph 3: Episodes watched per day on Netflix based on week, month, and year

Graph 3 shows that people are watching Netflix every single day of the week.

Frequency of activity in my netflix account per Month

View_month <- netflix_episode_perday %>%
  count(mesF)
View_monthView_month_plot <- View_month %>% 
  ggplot(aes(mesF, n)) +
  geom_col(fill = "blue") +
  coord_polar()  +
  theme_minimal() +
  theme(axis.title.x = element_blank(),
        axis.title.y = element_blank(),
        axis.text.y = element_blank(),
        axis.text.x = element_text(face = "bold"),
        plot.title = element_text(size = 18, face = "bold")) +
  ggtitle("Frequency of episodes watched", "Activity per month on Netflix") 
View_monthView_month_plot

Graph 4: Frequency of episodes watched activity per month

Graph 4 shows that viewers’ episode watch rate is high at the beginning of a year and in January due to the new year holiday time

Frequency of activity in my netflix account per Year

view_year <- netflix_episode_perday %>%
  count(añomes)
view_yearview_year_plot <- view_year %>% 
  ggplot(aes(añomes, n)) +
  geom_col(fill = "blue") +
  coord_polar()  +
  theme_minimal() +
  theme(axis.title.x = element_blank(),
        axis.title.y = element_blank(),
        axis.text.y = element_blank(),
        axis.text.x = element_text(face = "bold"),
        plot.title = element_text(size = 18, face = "bold")) +
  ggtitle("Frequency of episodes watched", "Activity by month of the year on Netflix")
view_yearview_year_plot

Graph 5: Frequency of episodes watched activity per month and year

It was observed that there were fewer frequent viewers in February from 2020  to 2022 and June of 2020-2021 in Graph 5.

From the data we can see that most users watch to love Netflix regularly. This is an interesting finding since it counters one of my previous assumptions about how people consume video content using Netflix these days.

References

NETFLIX. (n.d). Account. https://www.netflix.com/YourAccount