cryptography_ad_campaign_data_analysis
Esther_Wairimu_Kamau
1/10/2021
1.Defining the problem
The research problem in this case is to find out individuals that are likely to click on a blog advert based on their characteristics which include; Age Daily Time spent on site Area of residence Internet Usage Gender Country of residence
2.Metric of Success
The metric success of this project is to identify clients likely to click on the ad after performing intense data analysis(EDA).
3.Data Relevance
The data provided by the client is from the performance of a previous blog advert on the same website. The columns are as follows:
Daily Time Spent on the site-Integer
Age of the individual browsing-Integer
Area of residence Internet Usage
Gender of the browsing individual
Country of Residence
4.Understanding the Context
5.Experimental Design
- .Data Loading
- .Data cleaning for missing values and outliers
- .Exploratory Data Analysis
- .Conclusion-Detecting the trend in behaviour.
6.Data Loading and exploring
#### Importing our dataset
advertising =read.csv('http://bit.ly/IPAdvertisingData',header = TRUE, sep = ",",fileEncoding = "UTF-8-BOM")#### exploring the top of our data
head(advertising)## Daily.Time.Spent.on.Site Age Area.Income Daily.Internet.Usage
## 1 68.95 35 61833.90 256.09
## 2 80.23 31 68441.85 193.77
## 3 69.47 26 59785.94 236.50
## 4 74.15 29 54806.18 245.89
## 5 68.37 35 73889.99 225.58
## 6 59.99 23 59761.56 226.74
## Ad.Topic.Line City Male Country
## 1 Cloned 5thgeneration orchestration Wrightburgh 0 Tunisia
## 2 Monitored national standardization West Jodi 1 Nauru
## 3 Organic bottom-line service-desk Davidton 0 San Marino
## 4 Triple-buffered reciprocal time-frame West Terrifurt 1 Italy
## 5 Robust logistical utilization South Manuel 0 Iceland
## 6 Sharable client-driven software Jamieberg 1 Norway
## Timestamp Clicked.on.Ad
## 1 2016-03-27 00:53:11 0
## 2 2016-04-04 01:39:02 0
## 3 2016-03-13 20:35:42 0
## 4 2016-01-10 02:31:19 0
## 5 2016-06-03 03:36:18 0
## 6 2016-05-19 14:30:17 0#### exploring the bottom of our data
tail(advertising)## Daily.Time.Spent.on.Site Age Area.Income Daily.Internet.Usage
## 995 43.70 28 63126.96 173.01
## 996 72.97 30 71384.57 208.58
## 997 51.30 45 67782.17 134.42
## 998 51.63 51 42415.72 120.37
## 999 55.55 19 41920.79 187.95
## 1000 45.01 26 29875.80 178.35
## Ad.Topic.Line City Male
## 995 Front-line bifurcated ability Nicholasland 0
## 996 Fundamental modular algorithm Duffystad 1
## 997 Grass-roots cohesive monitoring New Darlene 1
## 998 Expanded intangible solution South Jessica 1
## 999 Proactive bandwidth-monitored policy West Steven 0
## 1000 Virtual 5thgeneration emulation Ronniemouth 0
## Country Timestamp Clicked.on.Ad
## 995 Mayotte 2016-04-04 03:57:48 1
## 996 Lebanon 2016-02-11 21:49:00 1
## 997 Bosnia and Herzegovina 2016-04-22 02:07:01 1
## 998 Mongolia 2016-02-01 17:24:57 1
## 999 Guatemala 2016-03-24 02:35:54 0
## 1000 Brazil 2016-06-03 21:43:21 1#### Checking the class of the object "advertising"
class(advertising)## [1] "data.frame"#### Checking the dimension of our dataset
dim(advertising)## [1] 1000 10#### Checking the structure of our data frame
str(advertising)## 'data.frame': 1000 obs. of 10 variables:
## $ Daily.Time.Spent.on.Site: num 69 80.2 69.5 74.2 68.4 ...
## $ Age : int 35 31 26 29 35 23 33 48 30 20 ...
## $ Area.Income : num 61834 68442 59786 54806 73890 ...
## $ Daily.Internet.Usage : num 256 194 236 246 226 ...
## $ Ad.Topic.Line : chr "Cloned 5thgeneration orchestration" "Monitored national standardization" "Organic bottom-line service-desk" "Triple-buffered reciprocal time-frame" ...
## $ City : chr "Wrightburgh" "West Jodi" "Davidton" "West Terrifurt" ...
## $ Male : int 0 1 0 1 0 1 0 1 1 1 ...
## $ Country : chr "Tunisia" "Nauru" "San Marino" "Italy" ...
## $ Timestamp : chr "2016-03-27 00:53:11" "2016-04-04 01:39:02" "2016-03-13 20:35:42" "2016-01-10 02:31:19" ...
## $ Clicked.on.Ad : int 0 0 0 0 0 0 0 1 0 0 ...We can observe that we have a mix of datatypes from intergers to strings
#### Getting the names of the columns we will be working with
colnames(advertising)## [1] "Daily.Time.Spent.on.Site" "Age"
## [3] "Area.Income" "Daily.Internet.Usage"
## [5] "Ad.Topic.Line" "City"
## [7] "Male" "Country"
## [9] "Timestamp" "Clicked.on.Ad"we can observe that our column names can all be changed to lowercase
7.Data cleaning
##### Checking for duplicated values in our data set
anyDuplicated(advertising)## [1] 0##### Checking if our dataset has any missing values
sum(is.na(advertising))## [1] 0### checking for missing values using case.complete function(just to confirm)
# The function complete.cases() returns a logical vector indicating which cases are complete.
# list rows of data that have missing values
advertising[!complete.cases(advertising),]## [1] Daily.Time.Spent.on.Site Age Area.Income
## [4] Daily.Internet.Usage Ad.Topic.Line City
## [7] Male Country Timestamp
## [10] Clicked.on.Ad
## <0 rows> (or 0-length row.names)### we rename the column names since they are too long
#we will be Using function rename
advertising=setnames(advertising, tolower(names(advertising[1:10])))
library(reshape)##
## Attaching package: 'reshape'## The following object is masked from 'package:tigerstats':
##
## tips## The following object is masked from 'package:Matrix':
##
## expand## The following object is masked from 'package:dplyr':
##
## rename## The following objects are masked from 'package:tidyr':
##
## expand, smiths## The following object is masked from 'package:data.table':
##
## meltadvertising <- rename(advertising, c(daily.time.spent.on.site="timespent"))
advertising <- rename(advertising, c(ad.topic.line="topic"))
advertising <- rename(advertising, c(daily.internet.usage="usage"))
advertising <- rename(advertising, c(clicked.on.ad ="clicked"))
advertising <- rename(advertising, c(timestamp="timestamp"))
advertising <- rename(advertising, c(area.income="income"))
advertising <- rename(advertising, c(male="gender"))### check if columns have been changed
head(advertising,n=3)## timespent age income usage topic city
## 1 68.95 35 61833.90 256.09 Cloned 5thgeneration orchestration Wrightburgh
## 2 80.23 31 68441.85 193.77 Monitored national standardization West Jodi
## 3 69.47 26 59785.94 236.50 Organic bottom-line service-desk Davidton
## gender country timestamp clicked
## 1 0 Tunisia 2016-03-27 00:53:11 0
## 2 1 Nauru 2016-04-04 01:39:02 0
## 3 0 San Marino 2016-03-13 20:35:42 0### checking for outliers, we only need the numerical columns
#first we get the numerical columns
nums <- unlist(lapply(advertising, is.numeric))
numcols <- advertising[ ,nums]
head(numcols,n=3)## timespent age income usage gender clicked
## 1 68.95 35 61833.90 256.09 0 0
## 2 80.23 31 68441.85 193.77 1 0
## 3 69.47 26 59785.94 236.50 0 0### checking for unique values
uniqueitems <- unique(advertising)
head(uniqueitems,n=3)## timespent age income usage topic city
## 1 68.95 35 61833.90 256.09 Cloned 5thgeneration orchestration Wrightburgh
## 2 80.23 31 68441.85 193.77 Monitored national standardization West Jodi
## 3 69.47 26 59785.94 236.50 Organic bottom-line service-desk Davidton
## gender country timestamp clicked
## 1 0 Tunisia 2016-03-27 00:53:11 0
## 2 1 Nauru 2016-04-04 01:39:02 0
## 3 0 San Marino 2016-03-13 20:35:42 0#### feature enginering the time/date
#we separate months,year and day each on its own
#library lubridate makes it easier for us to deal with dates
#install packages first then libraries
library(tidyr)
library(lubridate)##
## Attaching package: 'lubridate'## The following object is masked from 'package:reshape':
##
## stamp## The following objects are masked from 'package:data.table':
##
## hour, isoweek, mday, minute, month, quarter, second, wday, week,
## yday, year## The following objects are masked from 'package:base':
##
## date, intersect, setdiff, unionadvertising <- separate(advertising, timestamp, c("Year", "Month", "Day"))## Warning: Expected 3 pieces. Additional pieces discarded in 1000 rows [1, 2, 3,
## 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, ...].head(advertising,n=3)## timespent age income usage topic city
## 1 68.95 35 61833.90 256.09 Cloned 5thgeneration orchestration Wrightburgh
## 2 80.23 31 68441.85 193.77 Monitored national standardization West Jodi
## 3 69.47 26 59785.94 236.50 Organic bottom-line service-desk Davidton
## gender country Year Month Day clicked
## 1 0 Tunisia 2016 03 27 0
## 2 1 Nauru 2016 04 04 0
## 3 0 San Marino 2016 03 13 0#### Plotting the boxplot to visualize the outliers in the dataset
boxplot(numcols[,-1], horizontal=TRUE, main="Ad campaign outliers")
We observe that only income has any outliers,it wont affect the analysis so we countinue with the EDA.
8.Exploratory Data Analysis
Univariate Analysis
#For ease in analysis,we convert the data into a tibble REASONS why we use tibble dataframes
#never converts string as factor
#never changes the names of variables
#never create row names
library(tidyverse)
adv<-as_tibble(advertising)
head(adv,n=3)## # A tibble: 3 x 12
## timespent age income usage topic city gender country Year Month Day
## <dbl> <int> <dbl> <dbl> <chr> <chr> <int> <chr> <chr> <chr> <chr>
## 1 69.0 35 61834. 256. Clon~ Wrig~ 0 Tunisia 2016 03 27
## 2 80.2 31 68442. 194. Moni~ West~ 1 Nauru 2016 04 04
## 3 69.5 26 59786. 236. Orga~ Davi~ 0 San Ma~ 2016 03 13
## # ... with 1 more variable: clicked <int>Extracting Numerical tibble columns
#we define our tibble numerical dataframe
library(dplyr)
numt=adv %>% select_if(is.numeric)
head(numt,n=3)## # A tibble: 3 x 6
## timespent age income usage gender clicked
## <dbl> <int> <dbl> <dbl> <int> <int>
## 1 69.0 35 61834. 256. 0 0
## 2 80.2 31 68442. 194. 1 0
## 3 69.5 26 59786. 236. 0 0Extracting categorical tibble columns
Categoryt=adv %>% select_if(is.character)
head(Categoryt,n=3)## # A tibble: 3 x 6
## topic city country Year Month Day
## <chr> <chr> <chr> <chr> <chr> <chr>
## 1 Cloned 5thgeneration orchestration Wrightburgh Tunisia 2016 03 27
## 2 Monitored national standardization West Jodi Nauru 2016 04 04
## 3 Organic bottom-line service-desk Davidton San Marino 2016 03 13We first find the descriptive statistics of the numerical columns
summary(numt)## timespent age income usage
## Min. :32.60 Min. :19.00 Min. :13996 Min. :104.8
## 1st Qu.:51.36 1st Qu.:29.00 1st Qu.:47032 1st Qu.:138.8
## Median :68.22 Median :35.00 Median :57012 Median :183.1
## Mean :65.00 Mean :36.01 Mean :55000 Mean :180.0
## 3rd Qu.:78.55 3rd Qu.:42.00 3rd Qu.:65471 3rd Qu.:218.8
## Max. :91.43 Max. :61.00 Max. :79485 Max. :270.0
## gender clicked
## Min. :0.000 Min. :0.0
## 1st Qu.:0.000 1st Qu.:0.0
## Median :0.000 Median :0.5
## Mean :0.481 Mean :0.5
## 3rd Qu.:1.000 3rd Qu.:1.0
## Max. :1.000 Max. :1.0- We observe that mean of the age of individuals in our dataset is 36 with the oldest being 61.
- most individuals have an income of 55000 and with the lowest being 13996.
- the time spent online is mostly 1hr 5mins with the highest being 1hr 31mins.
- the cost of being online on hourly(65mins) rate is 180
- the mean of the page clicks is 0.5 meaning the clicks are equal to ‘no clicks’
Plotting Histograms for numerical columns
#par(mfrow = c(2, 2))
#hist(numt$timespent)
#hist(numt$age)
#hist(numt$income)
#hist(numt$usage)
#hist(numt$gender)
#hist(numt$clicked)numerical columns mode
#The mode is the value that appears most frequently in a data set
#Finding the mode of all numerical columns
#we start with age
v<-adv%>% pull(age)
getmode <- function(v) {
uniqv <- unique(v)
uniqv[which.max(tabulate(match(v, uniqv)))]
}
Age.Mode<-getmode(adv$age)
Age.Mode## [1] 31The age that appears most is 31years so most individuals who click on the page are in this age group
#we start with age
v2<-adv%>% pull(income)
getmode <- function(v) {
uniqv <- unique(v)
uniqv[which.max(tabulate(match(v, uniqv)))]
}
income.Mode<-getmode(adv$income)
income.Mode## [1] 61833.9We see that most individuals in dataset’s income is range of 60000 and above
#we start with age
v3<-adv%>% pull(timespent)
getmode <- function(v) {
uniqv <- unique(v)
uniqv[which.max(tabulate(match(v, uniqv)))]
}
time.Mode<-getmode(adv$timespent)
time.Mode## [1] 62.26We observe that most time spent that appears most times is 62 which means that our univariate plots were correct
#we start with age
v5<-adv%>% pull(usage)
getmode <- function(v) {
uniqv <- unique(v)
uniqv[which.max(tabulate(match(v, uniqv)))]
}
usage.Mode<-getmode(adv$usage)
usage.Mode## [1] 167.22most guys use 167 on every time they spend online.Which is almost same with the univariate plots.
### Plot frequency plots for categorical columns
#we start with country column
country <- Categoryt$country
Country_frequency<- table(country)
s<-desc(Country_frequency)
head(s,n=2)## country
## Afghanistan Albania
## -8 -7barplot(Country_frequency,col="Blue")
we observe that country that has most customers is Afghanistan followed albania as we can see in first console the second plot confirms it.
#secondly we tackle the city column
f2 <- Categoryt$city
f2_frequency<- table(f2)
g<-desc(f2_frequency)
head(g,n=3)## f2
## Adamsbury Adamside Adamsstad
## -1 -1 -1barplot(f2_frequency,col="Red")
we observe that williamsport city appears thrice more than most city column.It has too many unique values.
#
f3 <- Categoryt$Year
f3_frequency<- table(f3)
desc(f3_frequency)## f3
## 2016
## -1000barplot(f3_frequency,col="Yellow")
all observations were taken from 2016
f4 <- Categoryt$Month
f4_frequency<- table(f4)
desc(f4_frequency)## f4
## 01 02 03 04 05 06 07
## -147 -160 -156 -147 -147 -142 -101barplot(f4_frequency,col="Grey")
we observe that the month with highest traffic is February followed by march with january,April and may being the same.Also there is consistent traffic month on month.
f5 <- Categoryt$Day
f5_frequency<- table(f5)
head.matrix(f5_frequency)## f5
## 01 02 03 04 05 06
## 33 25 46 36 35 25barplot(f5_frequency,col="green")
we observe that no specific time of the month is there extra high traffic or extra low traffic is almost same all days.But on 31st we can notice is weirdly low.
f4 <- Categoryt$topic
f4_frequency<- table(f4)
head.matrix(f4_frequency)## f4
## Adaptive 24hour Graphic Interface Adaptive asynchronous attitude
## 1 1
## Adaptive context-sensitive application Adaptive contextually-based methodology
## 1 1
## Adaptive demand-driven knowledgebase Adaptive uniform capability
## 1 1barplot(f4_frequency,col="Blue",horiz=TRUE)
This means all topics have the same distribution they are too unique and none has counts than the other.
9.Bivariate Analysis
#clicks of individuals in our dataset month on month
ggplot(adv, aes(x = clicked,fill = Month)) + geom_bar(position = "stack")
we observe that the distrbution of individuals who clicked and the ones who didn’t is the same monthly.
#time spent online versus the income of individuals
geom_line()## geom_line: na.rm = FALSE, orientation = NA
## stat_identity: na.rm = FALSE
## position_identityggplot(data =adv,aes(x=timespent,y=usage))+
geom_line()
We observe that the more time on spends online the more the usage as we can see above
#time spent online versus the income of individuals
geom_line()## geom_line: na.rm = FALSE, orientation = NA
## stat_identity: na.rm = FALSE
## position_identityggplot(data =adv,aes(x=timespent,y=Day))+geom_line()
we observe that on a daily basis people spend time online on the page
# plot income changes by month, for each Gender
ggplot(adv, aes(x=Month, y = income)) +
geom_line(color="grey") +
geom_point(color="blue") +
facet_wrap(~gender) +
theme_minimal(base_size = 9) +
theme(axis.text.x = element_text(angle = 45,
hjust = 1)) +
labs(title = "Changes in income by Gender",
x = "Gender",
y = "income") 
we note that gender o has fewer individuals who earn below 20000 than gender 1 we also note that gender 0 and gender 1 almost have the same salaries over the months * in may and december there is partial disparity when it comes to the incomes gender o has more income earning individuals in those months than gender 1
# We check on the timespent versus the age and the click
qplot(x=age,data=adv,group=timespent,colour=clicked,bins=30)
we can observe that individuals as age decreases the clicks decrease but time spent in some ages like 30 increase alot.But from 38 to around 40 the time spent decreases but the clicks increase.
qplot(x=income,data=adv,group=timespent,colour=clicked,bins=30)
we can observe that the plot is skewed to right meaning that as income increases the more the more the time spent which also increases click
relationships between the target variable(clicked) and features
# Plot to show realtionship between clicked and income
qplot(income,
clicked,
data = adv,
geom = c("point", "smooth"),
alpha = I(1 / 5))## `geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
# Plot to show realtionship between clicked and income
qplot(age,
clicked,
data = adv,
geom = c("point", "smooth"),
alpha = I(1 / 5))## `geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
# Plot to show realtionship between clicked and income
qplot(usage,
clicked,
data = adv,
geom = c("point", "smooth"),
alpha = I(1 / 5))## `geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
# Plot to show realtionship between clicked and income
qplot(timespent,
clicked,
data = adv,
geom = c("point", "smooth"),
alpha = I(1 / 5))## `geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
10.Multivariate Analysis
#we look at the timespent considering usage groping by clicked or not clicked
# Color by groups; auto.key = TRUE to show legend
cloud(timespent ~ timespent * usage,
group = clicked, data = adv,
auto.key = TRUE)
we observe that most clicked spend alot time online and have high usage the purple cluster represents the clicked and blue not clicked.
#we look at time spent and usage versus the Gender
cloud(timespent ~ timespent * usage,
group = gender, data = adv,
auto.key = TRUE)
No Gender spends more time online than the other or has high usage than the other its the same
#we llok if Age affescts time spent online and page being clicked
cloud(timespent ~ timespent * age,
group = clicked, data = adv,
auto.key = TRUE)
We observe that As Age increases and time spent increases so does the click .But age seems to be clustered more in the middle when it comes to click which is purple.
library(corrplot)## corrplot 0.84 loaded# Compute a correlation matrix
corr <- round(cor((adv[0:4])),1)
corr## timespent age income usage
## timespent 1.0 -0.3 0.3 0.5
## age -0.3 1.0 -0.2 -0.4
## income 0.3 -0.2 1.0 0.3
## usage 0.5 -0.4 0.3 1.0corrplot(cor(corr), # Correlation matrix
method = "shade", # Correlation plot method
type = "full", # Correlation plot style (also "upper" and "lower")
diag = TRUE, # If TRUE (default), adds the diagonal
tl.col = "black", # Labels color
bg = "white", # Background color
title = "correalation matrix",# Main title
col = NULL) # Color palette
we observe that The income and Daily time spent on the site columns have a large positive correlation and so does the usage and timespent.Age has a very negative correlation with time spent
11.Recommendations
From our indepth Analysis we would advice our client to;
*come up with ad campaigns that lure young people especially the age group (28 to 30) who spent alot of time online.
- since gender is does not affect click she should still decide on her target market invest her resources.
*People who earn alot tend to be the biggest clickers but they dont spend alot of time online.Would recommend to client to come up with service flexible to any income earner since the usage is the same whether Wealthy or not.
12.Feature Importance
The dataset was appropriate. it contained no missing values and minimal outliers amongst the varaibles
Both univariate and Bivariate analysis revealed that the dataset is collinear, hence it can be analysed better by use of a classification algorithms