Customer_Data
Hammad Raza Khan
2024-04-29
#Adding the Dataset
library(readr)
customer_data <- read_csv("C:/Users/hamma/Downloads/customer_data.csv")## New names:
## Rows: 202 Columns: 16
## ── Column specification
## ──────────────────────────────────────────────────────── Delimiter: "," chr
## (7): gender, education, region, loyalty_status, purchase_frequency, prod... dbl
## (6): id, age, income, purchase_amount, promotion_usage, satisfaction_score lgl
## (3): ...13, ...14, ...15
## ℹ Use `spec()` to retrieve the full column specification for this data. ℹ
## Specify the column types or set `show_col_types = FALSE` to quiet this message.
## • `` -> `...13`
## • `` -> `...14`
## • `` -> `...15`
## • `` -> `...16`View(customer_data)Topic / Dataset = “CUSTOMER_DATA”
1. PRE-PROCESSING OF DATA
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, unionlibrary(tidyverse)## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ forcats 1.0.0 ✔ stringr 1.5.1
## ✔ ggplot2 3.5.1 ✔ tibble 3.2.1
## ✔ lubridate 1.9.3 ✔ tidyr 1.3.1
## ✔ purrr 1.0.2## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errorslibrary(tidyr)
library(ggplot2)
library(knitr)
# DATA PRE-PROCESSING
# deleting null values rows
customer_data = customer_data[-200, ]
view(customer_data)ANALYSIS OF DATA
ANALYSIS - 1
#ANALYSIS 1
# Age Distribution: What is the distribution of ages among customers?
ggplot(data = customer_data, aes(x = age)) +
geom_histogram(binwidth = 1) +
theme(plot.background = element_rect(fill = "skyblue", color = "red")) +
labs(title = "Age Frequency",
x = "ages",
y = "Frequency") ## Warning: Removed 2 rows containing non-finite outside the scale range
## (`stat_bin()`).
##Since there are outliners, they should be taken care of -
customer_data %>%
group_by(age) %>%
summarise(n()) ## # A tibble: 23 × 2
## age `n()`
## <dbl> <int>
## 1 20 1
## 2 21 2
## 3 22 1
## 4 23 8
## 5 24 7
## 6 25 13
## 7 26 7
## 8 27 15
## 9 28 21
## 10 29 21
## # ℹ 13 more rows#After excluding Out-liners
ggplot(data = customer_data, aes(x = age)) +
geom_histogram(binwidth = 1) +
xlim(23, 40) +
theme(plot.background = element_rect(fill = "skyblue", color = "red")) +
labs(title = "Age Frequency",
x = "ages",
y = "Frequency") ## Warning: Removed 7 rows containing non-finite outside the scale range
## (`stat_bin()`).## Warning: Removed 2 rows containing missing values or values outside the scale range
## (`geom_bar()`).
##OUTCOME: After using the group_by() function, we could easily define the OUTLINERS of the data and remove them from the histogram graph. The ages between 25 and 30 tend to have the highest frequency in the dataset.
#ANALYSIS 2
# Purchase Amount by Education Level: How does the education level of customers affect
# their purchase amount?
# Visualization Tool Used : BOX PLOT
ggplot(customer_data, aes(x = education, y = purchase_amount)) +
geom_boxplot(fill = "skyblue", color = "black") +
labs(title = "Purchase Amount by Education Level",
x = "Education Level", y = "Purchase Amount") +
theme_minimal()## Warning: Removed 2 rows containing non-finite outside the scale range
## (`stat_boxplot()`).
## OUTCOME: Since, the box-plot is divided into 4 Quarters, we can
analyze the Quarters for each of the category as follows Bachelors
College High School Masters 1st Quarter: 2500-5000 | 2500-5100 |
2600-4900 | 2500-6000 | 2nd Quarter: 5000-10,200 | 5100-9000 | 4900-8000
| 6000-9000 | 3rd Quarter: 10,200-11,500 | 9000-12000 | 8000-12600 |
9000-13000 | 4th Quarter: 11,500-22,000 | 12000-21000 | 12600-23000 |
13000-17000 | Median : 10,500 | 9000 | 8000 | 9000 |
Note: (These are approx values)
##Analysis-3 ##Now, Using Box-Plot for the same info to get a better idea of the data.
ggplot(customer_data, aes(x = education, y = mean(customer_data$purchase_amount))) +
geom_bar(stat = "identity", fill = "skyblue", color = "black") +
labs(title = "Simple Bar Plot", x = "Category", y = "Value") +
theme_minimal()## Warning: Use of `customer_data$purchase_amount` is discouraged.
## ℹ Use `purchase_amount` instead.## Warning: Position guide is perpendicular to the intended axis.
## ℹ Did you mean to specify a different guide `position`?## Warning: Removed 201 rows containing missing values or values outside the scale range
## (`geom_bar()`).
#ANALYSIS - 4 ##Income vs. Purchase Amount: Is there a relationship between income and purchase amount?
# Visualization Tool Used: scatter plot
ggplot(data = customer_data, aes(x = income, y= purchase_amount)) +
geom_point() +
labs(title = "Income vs. Purchase Amount",
x = "Income",
y = "Purchase Amount")## Warning: Removed 2 rows containing missing values or values outside the scale range
## (`geom_point()`).
OUTCOME: Since the dots can be seen in a linear direction; If the variables correlate they will fall along a line or curve. The stronger the correlation the tighter the data points will follow the line or curve. Therefore, these columns are correlatable… dependent variables (y-axis): purchase_amount independent variables(x-axis): income CORRELATION: POSITIVE
#Analysis-5 # Income Distribution: What is the distribution of income among the customers?
ggplot(data = customer_data, aes(x = income)) +
geom_histogram(binwidth = 5000) +
theme(plot.background = element_rect(fill = "skyblue", color = "red")) +
labs(title = "Income Frequency",
x = "Income",
y = "Frequency")## Warning: Removed 2 rows containing non-finite outside the scale range
## (`stat_bin()`).
OUTCOME: Certainly! When you execute this code, you’ll see a graphical
representation that looks like a bar chart. The horizontal axis, labeled
“Income,” shows different income ranges, while the vertical axis,
labeled “Frequency,” represents how many customers have incomes within
each range. Each bar on the chart corresponds to a particular income
range, and its height indicates the number of customers falling within
that range. This visualization helps you understand the distribution of
income among the customers in the dataset.
#Analaysis-6
# Income vs. Purchase Amount by Gender: How does the relationship between
# income and purchase amount vary between different genders?
#Visual Tool Used: Scatter Plot
ggplot(customer_data, aes(x = income, y = purchase_amount, color = gender, shape = gender)) +
geom_point() +
labs(title = "Relationship Between Income and Purchase Amount by Gender",
x = "Income", y = "Purchase Amount",
color = "Gender", shape = "Gender") +
scale_color_manual(values = c("blue", "pink")) +
scale_shape_manual(values = c(16, 17)) +
theme_minimal()## Warning: Removed 2 rows containing missing values or values outside the scale range
## (`geom_point()`).
OUTCOME: Since the dots can be seen in a linear direction; If the
variables correlate they will fall along a line or curve. The stronger
the correlation the tighter the data points will follow the line or
curve. Therefore, these columns are correlatable… dependent variables
(y-axis): purchase_amount independent variables(x-axis): income
CORRELATION: POSITIVE
#Analysis-7
# Calculate proportion of promotion usage for each loyalty status
# Create a table of counts for each loyalty status and promotion usage
table_data = table(customer_data$loyalty_status,
ifelse(customer_data$promotion_usage >= 0.5, "Used", "Not Used"))
# Convert counts to proportions
prop_data = prop.table(table_data, margin = 1)
# Create a stacked bar plot
barplot(prop_data,
beside = TRUE,
main = "Promotion Usage by Loyalty Status",
xlab = "Loyalty Status",
ylab = "Proportion of Promotion Usage",
col = c("gold", "lightgreen","grey"),
legend = rownames(prop_data),
args.legend = list(x = "topright")
)
OUTCOME: This code will generate a grouped bar chart where each bar
represents a different loyalty status level, and within each bar, there
are stacked bars representing the proportion of promotion usage. Adjust
the data and plot settings according to your actual data and
preferences.
#ANALYSIS-8
# Relationship Between Age, Income, and Purchase Amount:
# How do age, income, and purchase amount relate to each other in the dataset?
data <- data.frame(Age = customer_data$age, Income = customer_data$income, Purchase_Amount = customer_data$purchase_amount)
# Create a pair plot
ggplot(data, aes(x = Age, y = Income)) +
geom_point() +
geom_smooth(method = "lm", se = FALSE, color = "blue") +
labs(title = "Pair Plot of Age vs. Income") +
theme_minimal()## `geom_smooth()` using formula = 'y ~ x'## Warning: Removed 2 rows containing non-finite outside the scale range
## (`stat_smooth()`).## Warning: Removed 2 rows containing missing values or values outside the scale range
## (`geom_point()`).
ggplot(data, aes(x = Age, y = Purchase_Amount)) +
geom_point() +
geom_smooth(method = "lm", se = FALSE, color = "green") +
labs(title = "Pair Plot of Age vs. Purchase Amount") +
theme_minimal()## `geom_smooth()` using formula = 'y ~ x'## Warning: Removed 2 rows containing non-finite outside the scale range (`stat_smooth()`).
## Removed 2 rows containing missing values or values outside the scale range
## (`geom_point()`).
ggplot(data, aes(x = Income, y = Purchase_Amount)) +
geom_point() +
geom_smooth(method = "lm", se = FALSE, color = "red") +
labs(title = "Pair Plot of Income vs. Purchase Amount") +
theme_minimal()## `geom_smooth()` using formula = 'y ~ x'## Warning: Removed 2 rows containing non-finite outside the scale range (`stat_smooth()`).
## Removed 2 rows containing missing values or values outside the scale range
## (`geom_point()`).
Correlation : Linear correlation found between Income and Purchase Amount
#ANALYSIS-9
# Predicting Satisfaction Income on Purhcase_Amount: Can we predict the satisfaction score of
# customers based on their age?
# VU: A single regression analysis with age as the independent variable and
# satisfaction score as the dependent variable. Interpret the regression coefficients
# and assess the predictive power of the model.
# Create a dataframe with age and satisfaction score
data <- data.frame(Age = customer_data$age, Satisfaction_Score = customer_data$satisfaction_score)
# Perform linear regression
model <- lm(Satisfaction_Score ~ Age, data = data)
# Print the summary of the regression model
summary(model)##
## Call:
## lm(formula = Satisfaction_Score ~ Age, data = data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.12383 -0.83034 0.02291 0.87617 2.36532
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.50965 0.54166 6.479 7.21e-10 ***
## Age 0.04891 0.01791 2.732 0.00687 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.067 on 197 degrees of freedom
## (2 observations deleted due to missingness)
## Multiple R-squared: 0.0365, Adjusted R-squared: 0.03161
## F-statistic: 7.463 on 1 and 197 DF, p-value: 0.00687# Scatter plot with regression line
ggplot(data, aes(x = Age, y = Satisfaction_Score)) +
geom_point() + # Scatter plot
geom_smooth(method = "lm", se = FALSE, color = "blue") + # Regression line
labs(title = "Scatter Plot of Age vs. Satisfaction Score with Regression Line",
x = "Age", y = "Satisfaction Score") + # Axes labels
theme_minimal()## `geom_smooth()` using formula = 'y ~ x'## Warning: Removed 2 rows containing non-finite outside the scale range
## (`stat_smooth()`).## Warning: Removed 2 rows containing missing values or values outside the scale range
## (`geom_point()`).
#Analysis-10 It was a poor representation; so I decided to represent it
in side-by-side bar graph instead.
data <- data.frame(
Age = customer_data$age,
Satisfaction_score = customer_data$satisfaction_score)
model <- lm(Satisfaction_score ~ Age, data = data)
ggplot(data, aes(x = Age)) +
geom_bar(aes(fill = factor(Satisfaction_score)), position = "dodge") +
labs(title = "Side-by-Side Bar Chart of Age and Satisfaction Score",
x = "Age", y = "Frequency") +
theme_minimal()+
coord_flip()## Warning: Removed 2 rows containing non-finite outside the scale range
## (`stat_count()`).
Prediction: Customers between the range of 35-40 tend to have the
highest satisfaction of their product.
#ANALYSIS-11
#Correlation Between Purchase Frequency, Product Category, and Income:
# Is there a visual correlation between purchase frequency, product category, and income?
# VU: Clustered bar chart to visualize the relationship between these variables.
ggplot(customer_data, aes(x = customer_data$product_category, fill = customer_data$income)) +
geom_bar(position = "dodge") +
facet_wrap(~ customer_data$gender) +
labs(title = "Clustered Bar Chart of Purchase Frequency by Product Category and Gender",
x = "Product Category", y = "Frequency",
fill = "Income") +
theme_minimal()## Warning: Use of `customer_data$product_category` is discouraged.
## ℹ Use `product_category` instead.## Warning: Use of `customer_data$income` is discouraged.
## ℹ Use `income` instead.## Warning: The following aesthetics were dropped during statistical transformation: fill.
## ℹ This can happen when ggplot fails to infer the correct grouping structure in
## the data.
## ℹ Did you forget to specify a `group` aesthetic or to convert a numerical
## variable into a factor?
## The following aesthetics were dropped during statistical transformation: fill.
## ℹ This can happen when ggplot fails to infer the correct grouping structure in
## the data.
## ℹ Did you forget to specify a `group` aesthetic or to convert a numerical
## variable into a factor?
Expected Outcome: This code creates a clustered bar chart using ggplot2
to visualize the relationship between purchase frequency, product
category, income, and gender in the customer_data dataset.y.
Additionally, the plot is faceted by gender, allowing for separate
visualizations of purchase frequency for each gender. This visualization
helps to explore how purchase frequency varies across different product
categories, income levels, and genders in the dataset. Outcome
Highlights: Female tend to have higher interest in Electronics then
Male; while Male tend to have higher interest in beauty !
#ANALYSIS-12
#Comparison of Purchase Frequency Across Different Age Groups and Product Categories:
#How does purchase frequency vary across different age groups and product categories?
# VU: A grouped bar chart
purchase_data <- data.frame(
Age_group = customer_data$age,
Product_category = customer_data$product_category,
Purchase_frequency = customer_data$purchase_frequency)
# Created a grouped bar chart
ggplot(purchase_data, aes(x = Age_group, y = Purchase_frequency, fill = Product_category)) +
geom_bar(stat = "identity", position = "dodge") +
labs(title = "Purchase Frequency Across Age Groups and Product Categories",
x = "Age Group", y = "Purchase Frequency",
fill = "Product Category") +
theme_minimal()## Warning: Removed 2 rows containing missing values or values outside the scale range
## (`geom_bar()`).
Outcome: a grouped bar chart showing the purchase frequency for
different product categories across different age groups. Each group of
bars represents a specific age group, and within each group, bars of
different colors represent different product categories. This
visualization helps to explore how purchase frequency varies across age
groups and product categories in the dataset. Outcome: Highlights: The
age group of age higher than 40 is more interested in Books. Electronics
are the most often bought products.
#ANALYSIS-13
# Multivariate Analysis of Customer Behavior: How do multiple variables
# (e.g., age, income, education level, and region) collectively influence
# purchasing behavior?
# VU: Multidimensional scatter plot or ternary plot to visualize the relationships between these variables and their impact on purchasing behavior.
library(ggplot2)
cu <- data.frame(
Income = customer_data$income,
Education_Qualification = customer_data$education)
# Count the number of observations for each combination of income and education qualification
income_education_count <- table(customer_data$income, customer_data$education)
# Convert the table to a data frame
income_education_count <- as.data.frame(income_education_count)
colnames(income_education_count) <- c("Income", "Education_Qualification", "Count")
# Create a pie chart
ggplot(income_education_count, aes(x = "", y = Count, fill = Education_Qualification)) +
geom_bar(stat = "identity", width = 1) +
coord_polar("y", start = 0) +
labs(title = "Distribution of Income Across Education Qualifications",
fill = "Education Qualification") +
theme_minimal() +
theme(legend.position = "right")
Expected Outcome: . Each segment of the pie chart represents a specific
education qualification, and the size of each segment corresponds to the
count of observations for that qualification. This visualization helps
to understand how income is distributed among individuals with different
education qualifications. Outcome Highlights: College Degree Education
tend to have the Income in the customer data; while the Masters degree
customers are having the lowest ones.
#ANALYSIS-14
# Perform ANOVA
anova_result <- aov(customer_data$purchase_amount ~ customer_data$education * customer_data$region, data = customer_data)
# Summary of ANOVA
summary(anova_result)## Df Sum Sq Mean Sq F value
## customer_data$education 3 3.890e+07 12966775 0.449
## customer_data$region 3 2.293e+07 7643301 0.265
## customer_data$education:customer_data$region 9 1.259e+08 13984331 0.485
## Residuals 183 5.281e+09 28857971
## Pr(>F)
## customer_data$education 0.718
## customer_data$region 0.851
## customer_data$education:customer_data$region 0.884
## Residuals
## 2 observations deleted due to missingness# Visualize the results with boxplots
ggplot(customer_data, aes(x = education, y = purchase_amount, fill = region)) +
geom_boxplot() +
labs(title = "Purchase Amount by Education Level and Region",
x = "Education Level", y = "Purchase Amount",
fill = "Region") +
theme_minimal()## Warning: Removed 2 rows containing non-finite outside the scale range
## (`stat_boxplot()`).
Expected Outcome: A visualization of purchase amount by education level
and region using boxplots. This allows you to explore the variation in
purchase amount across different education levels and regions, and
assess whether there are significant differences in purchase amount
between groups.
Pie chart showing the distribution of income across different education qualifications in the dataset. Each slice of the pie represents a specific education qualification, and the size of the slice corresponds to the proportion of individuals with that education qualification in each income category.
#ANALYSIS -15
# Load required libraries
library(ggplot2)
# Calculate Pearson correlation coefficient
correlation_coefficient <- cor(customer_data$age, customer_data$satisfaction_score)
# Print the correlation coefficient
print(paste("Pearson correlation coefficient:", correlation_coefficient))## [1] "Pearson correlation coefficient: NA"# Create scatter plot
ggplot(customer_data, aes(x = age, y = satisfaction_score)) +
geom_point() +
labs(title = "Age vs. Satisfaction Score",
x = "Age", y = "Satisfaction Score") +
theme_minimal()## Warning: Removed 2 rows containing missing values or values outside the scale range
## (`geom_point()`).
This is to show what a poor analysis of scatter plot would look like.
When the dots are supposed to be in a linear regression; they are
scrambled here and there instead.
#ANALYSIS-16
# Load required libraries
library(ggplot2)
# Calculate average income for males and females separately
average_income <- aggregate(income ~ gender, data = customer_data, FUN = mean)
# Create pie chart
ggplot(average_income, aes(x = "", y = income, fill = gender)) +
geom_bar(stat = "identity", width = 1) +
coord_polar("y", start = 0) +
labs(title = "Average Income by Gender",
x = NULL, y = NULL,
fill = "Gender") +
theme_void() +
theme(legend.position = "right")
#ANALYSIS-17
# Load required libraries
library(ggplot2)
# Create box plot
ggplot(customer_data, aes(x = gender, y = income, fill = gender)) +
geom_boxplot() +
labs(title = "Income Distribution by Gender",
x = "Gender", y = "Income",
fill = "Gender") +
theme_minimal()## Warning: Removed 2 rows containing non-finite outside the scale range
## (`stat_boxplot()`).