# E-Commerce Data Project
options(repos = c(CRAN = "https://cloud.r-project.org/"))
# set up my working directory (you need to replace with your own)
setwd("/Users/mag/Desktop/Project 2")
# STEP 1: Load Required Libraries
install.packages("data.table")
##
## The downloaded binary packages are in
## /var/folders/zg/cx_xk6ks2q95n18cjvn4wb140000gn/T//RtmpSpgsy2/downloaded_packages
install.packages("cluster")
##
## The downloaded binary packages are in
## /var/folders/zg/cx_xk6ks2q95n18cjvn4wb140000gn/T//RtmpSpgsy2/downloaded_packages
install.packages("factoextra")
##
## The downloaded binary packages are in
## /var/folders/zg/cx_xk6ks2q95n18cjvn4wb140000gn/T//RtmpSpgsy2/downloaded_packages
install.packages("plotly")
##
## The downloaded binary packages are in
## /var/folders/zg/cx_xk6ks2q95n18cjvn4wb140000gn/T//RtmpSpgsy2/downloaded_packages
library(ggplot2) # For data visualization
library(dplyr) # For data manipulation
##
## 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(data.table) # For data handling
##
## Attaching package: 'data.table'
## The following objects are masked from 'package:dplyr':
##
## between, first, last
library(cluster) # For clustering analysis
library(factoextra) # For clustering visualization
## Welcome! Want to learn more? See two factoextra-related books at https://goo.gl/ve3WBa
library(plotly) # For interactive visualization
##
## 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
# STEP 2: Load & Explore the dataset (retail.segmentation.csv)
## Read the dataset (Assuming it's in your working directory)
retail_segmentation <-read.csv("retail_segmentation.csv")
## Check the first few rows
head(retail_segmentation)
## Cust_No avg_order_size avg_order_freq crossbuy multichannel per_sale tenure
## 1 1 23.400000 2.2222222 3 2 0.00000000 3
## 2 2 34.260377 6.6250000 7 2 0.11111111 35
## 3 3 43.575641 4.8750000 5 2 0.07407407 12
## 4 4 26.316667 0.9000000 4 2 0.25000000 9
## 5 5 8.269231 1.0833333 3 1 0.50000000 40
## 6 6 21.500000 0.2222222 1 2 0.00000000 7
## return_rate married own_home household_size loyalty_card income age
## 1 0.1175214 1 1 1 1 35 47
## 2 0.2818684 1 1 3 1 140 70
## 3 0.2741769 1 0 4 0 35 21
## 4 0.1435508 0 0 1 1 35 62
## 5 0.0000000 0 0 2 0 140 21
## 6 0.0000000 0 1 1 1 80 21
## avg_mktg_cnt zip_code
## 1 56.000000 21230
## 2 14.914286 22301
## 3 20.083333 19002
## 4 8.222222 22304
## 5 1.350000 20124
## 6 2.714286 22033
## Check structure of data
str(retail_segmentation)
## 'data.frame': 2000 obs. of 16 variables:
## $ Cust_No : int 1 2 3 4 5 6 7 8 9 10 ...
## $ avg_order_size: num 23.4 34.26 43.58 26.32 8.27 ...
## $ avg_order_freq: num 2.22 6.62 4.88 0.9 1.08 ...
## $ crossbuy : int 3 7 5 4 3 1 2 3 1 1 ...
## $ multichannel : int 2 2 2 2 1 2 1 2 1 1 ...
## $ per_sale : num 0 0.1111 0.0741 0.25 0.5 ...
## $ tenure : int 3 35 12 9 40 7 8 17 14 3 ...
## $ return_rate : num 0.118 0.282 0.274 0.144 0 ...
## $ married : int 1 1 1 0 0 0 1 0 0 0 ...
## $ own_home : int 1 1 0 0 0 1 0 1 1 1 ...
## $ household_size: int 1 3 4 1 2 1 1 1 2 8 ...
## $ loyalty_card : int 1 1 0 1 0 1 0 1 1 1 ...
## $ income : int 35 140 35 35 140 80 70 35 35 35 ...
## $ age : int 47 70 21 62 21 21 86 70 57 21 ...
## $ avg_mktg_cnt : num 56 14.91 20.08 8.22 1.35 ...
## $ zip_code : int 21230 22301 19002 22304 20124 22033 8757 8109 21122 21208 ...
## Summary statistics
summary(retail_segmentation)
## Cust_No avg_order_size avg_order_freq crossbuy
## Min. : 1.0 Min. : 1.833 Min. : 0.02778 Min. :1.000
## 1st Qu.: 500.8 1st Qu.: 23.157 1st Qu.: 0.30769 1st Qu.:1.000
## Median :1000.5 Median : 30.790 Median : 0.76923 Median :2.000
## Mean :1000.5 Mean : 35.373 Mean : 1.55640 Mean :2.608
## 3rd Qu.:1500.2 3rd Qu.: 40.959 3rd Qu.: 1.90584 3rd Qu.:4.000
## Max. :2000.0 Max. :528.250 Max. :31.87500 Max. :7.000
## multichannel per_sale tenure return_rate
## Min. :1.000 Min. :0.0000 Min. : 1.00 Min. :0.00000
## 1st Qu.:1.000 1st Qu.:0.0000 1st Qu.: 4.00 1st Qu.:0.00000
## Median :1.000 Median :0.0000 Median :10.00 Median :0.01947
## Mean :1.557 Mean :0.1033 Mean :14.12 Mean :0.17671
## 3rd Qu.:2.000 3rd Qu.:0.1400 3rd Qu.:20.00 3rd Qu.:0.24560
## Max. :3.000 Max. :1.0000 Max. :40.00 Max. :6.90909
## married own_home household_size loyalty_card
## Min. :0.0000 Min. :0.000 Min. :1.000 Min. :0.0000
## 1st Qu.:0.0000 1st Qu.:0.000 1st Qu.:1.000 1st Qu.:0.0000
## Median :0.0000 Median :1.000 Median :2.000 Median :1.0000
## Mean :0.4635 Mean :0.568 Mean :2.869 Mean :0.6185
## 3rd Qu.:1.0000 3rd Qu.:1.000 3rd Qu.:4.000 3rd Qu.:1.0000
## Max. :1.0000 Max. :1.000 Max. :8.000 Max. :1.0000
## income age avg_mktg_cnt zip_code
## Min. : 35.00 Min. :21.00 Min. : 0.00 Min. : 7726
## 1st Qu.: 35.00 1st Qu.:21.00 1st Qu.: 4.00 1st Qu.:19010
## Median : 60.00 Median :37.00 Median : 7.75 Median :20854
## Mean : 75.79 Mean :42.93 Mean : 12.94 Mean :18326
## 3rd Qu.:110.00 3rd Qu.:61.25 3rd Qu.: 15.00 3rd Qu.:21776
## Max. :175.00 Max. :99.00 Max. :297.00 Max. :24060
# STEP 3: Preprocessing the Data (Manipulation)
## Compute sales and profit. In our dataset sales and profit
## are defined as follows: sales = avg_order_freg X avg_prder_size
## profit = 0.52 X sales – 0.75 X avg_mktg_cnt
retail_segmentation <- retail_segmentation %>%
mutate(sales = avg_order_freq * avg_order_size,
profit = 0.52 * sales - 0.75 * avg_mktg_cnt)
### View the updated dataframe
View(retail_segmentation) ## see the new columns created.
#STEP 4: Select variables for clustering analysis
## In your work, you can replace one or two variables from the list.
cluster_data <-retail_segmentation %>% select(tenure, own_home,
married, household_size, income, sales, profit)
# STEP 5: Determine the Optional Number of Clusters (Elbow Method)
## We can use the elbow method to determine the optimal number of
## clusters. The elbow method plots the within-cluster sum of squares
## (WCSS) against the number of clusters.
## The "elbow" point indicates the optimal number of clusters.
fviz_nbclust(cluster_data, kmeans, k.max = 20, method = "wss")
## Notes on the Elbow chart.
### Look for the "elbow point" where the WCSS (within sum of squares)
### stop decreasing significantly.
###For our example, we will choose 6 clusters (judgmental from
### the "elbow chart" point)
# STEP 6: Perform k-means clustering (Assuming k=6 from elbow point)
set.seed (123) ## Ensure reproducibility in random number generation.
kmeans_model <- kmeans(cluster_data, centers = 6)
## View model results
kmeans_model
## K-means clustering with 6 clusters of sizes 28, 93, 571, 130, 912, 266
##
## Cluster means:
## tenure own_home married household_size income sales profit
## 1 17.82143 0.5000000 0.4285714 2.785714 96.07143 433.91012 204.7848768
## 2 15.16129 0.6129032 0.4516129 3.290323 60.48387 205.77286 89.9889359
## 3 16.17863 0.5183888 0.5271454 2.760070 127.76708 22.36598 4.3885284
## 4 18.75385 0.4000000 0.6230769 2.569231 125.34615 111.75165 44.1802056
## 5 11.61294 0.6250000 0.3980263 2.968202 45.78947 17.75938 0.5171155
## 6 15.28947 0.5526316 0.4812030 2.770677 46.05263 85.81361 32.0590063
##
## Clustering vector:
## [1] 5 2 2 5 3 5 5 5 5 5 5 3 5 4 6 5 5 3 5 3 5 3 5 1 3 2 3 3 3 3 3 5 5 6 6 5 5
## [38] 6 6 3 6 4 5 5 5 5 5 6 4 5 5 5 5 5 3 5 3 5 3 5 1 3 5 2 5 5 5 5 3 5 5 6 6 6
## [75] 4 6 4 5 6 2 3 3 5 6 3 5 3 2 5 5 5 3 5 5 2 6 3 5 2 5 5 6 5 2 5 3 4 5 5 6 3
## [112] 5 5 5 5 5 5 5 3 5 5 3 5 5 5 3 2 3 5 6 2 4 6 6 3 4 5 5 5 5 5 5 5 5 3 5 3 5
## [149] 3 3 5 3 3 6 3 3 5 5 5 5 3 4 5 5 3 5 3 5 6 4 5 3 3 5 3 5 5 5 3 5 2 2 3 5 5
## [186] 5 5 3 6 4 3 3 2 5 5 5 2 4 6 5 4 5 3 5 3 5 4 5 5 5 5 5 5 3 2 5 3 5 3 3 5 3
## [223] 5 3 3 5 5 5 5 2 5 6 3 6 5 5 5 5 2 5 5 5 4 3 6 3 5 4 2 5 2 5 5 5 5 3 5 1 5
## [260] 6 5 6 3 5 3 3 5 5 5 3 3 6 3 3 5 3 3 6 6 4 5 4 5 3 3 6 5 3 5 3 6 6 5 5 3 6
## [297] 6 3 5 5 3 6 3 5 5 5 6 4 3 5 3 5 6 5 6 5 4 5 3 5 6 5 5 2 5 5 5 5 3 2 4 3 6
## [334] 5 5 3 5 3 3 3 3 3 3 5 5 4 3 3 5 5 5 5 3 5 6 5 6 1 5 2 6 3 4 6 6 5 5 3 3 6
## [371] 5 3 3 3 3 1 6 6 3 5 5 2 5 2 5 3 5 6 3 4 5 5 2 3 3 5 6 4 4 3 5 5 5 5 5 3 6
## [408] 6 2 3 3 5 5 5 6 5 5 3 6 3 2 5 5 5 5 5 5 5 4 6 3 5 6 2 5 5 2 5 3 4 6 5 2 5
## [445] 3 3 5 5 3 5 6 1 3 5 5 5 3 5 3 3 5 5 5 2 5 3 5 2 5 5 3 3 5 3 3 3 5 5 5 5 2
## [482] 3 5 5 3 5 5 3 5 3 5 5 3 3 6 3 4 5 3 2 5 5 3 5 3 6 6 3 3 3 3 5 3 6 5 3 6 5
## [519] 5 6 5 5 5 3 3 3 6 6 3 5 4 5 5 3 3 3 3 3 4 5 3 3 3 5 4 5 6 4 6 6 3 5 2 5 1
## [556] 5 5 5 5 3 5 5 4 6 6 3 5 3 3 4 5 6 3 2 6 3 5 3 4 6 3 6 4 4 3 5 1 5 5 2 5 3
## [593] 6 6 5 3 5 3 3 6 3 3 5 6 3 3 5 5 5 2 6 3 5 1 5 5 5 5 4 3 5 3 6 5 5 5 5 3 3
## [630] 5 5 2 5 5 5 5 3 5 5 3 5 3 5 5 5 5 5 6 6 3 6 6 6 3 1 2 2 5 5 5 5 5 4 3 5 5
## [667] 5 4 6 5 6 3 6 6 2 3 5 4 3 5 5 5 5 4 4 5 3 3 5 3 5 5 6 6 5 5 3 4 5 3 2 5 3
## [704] 3 4 3 3 6 5 5 1 6 2 4 1 5 3 5 5 3 5 6 6 6 5 5 5 5 5 5 4 3 5 3 6 5 3 5 6 6
## [741] 5 5 5 5 5 5 3 5 5 4 3 5 4 4 5 5 3 5 4 5 6 5 5 4 3 6 6 3 3 3 3 6 4 5 4 5 5
## [778] 5 3 5 6 1 6 5 2 5 6 5 5 5 3 3 3 6 5 5 6 3 5 5 6 3 6 6 3 3 5 5 3 3 5 5 5 4
## [815] 5 2 3 5 3 5 3 5 4 5 6 4 3 5 4 5 5 3 6 4 3 6 3 3 5 5 3 5 3 5 5 5 5 5 5 5 3
## [852] 5 6 3 3 5 5 3 5 5 3 3 3 5 3 3 5 2 5 5 6 5 5 3 6 6 3 5 3 5 5 5 3 5 5 3 3 5
## [889] 3 6 5 6 5 5 3 4 5 3 3 6 4 5 4 3 3 3 5 5 5 5 4 5 3 5 5 5 6 5 5 3 5 5 3 6 6
## [926] 3 3 5 6 5 5 5 4 5 5 5 5 5 5 5 3 5 3 5 6 2 5 4 6 5 5 5 6 4 3 3 6 5 5 5 4 3
## [963] 6 5 5 5 3 6 6 6 5 3 5 5 5 5 3 5 5 2 6 6 5 3 5 3 3 5 4 5 5 5 5 5 4 3 5 3 3
## [1000] 3 3 3 6 6 6 3 6 3 5 5 3 3 5 3 5 2 3 3 3 5 6 5 5 3 5 5 3 6 3 1 3 4 3 5 5 2
## [1037] 5 5 5 5 5 6 5 3 5 5 5 5 5 3 4 3 5 4 5 5 5 4 6 5 3 6 3 5 3 2 5 5 5 5 5 3 3
## [1074] 3 5 3 6 3 6 6 5 5 6 2 3 6 3 5 5 6 5 6 3 3 5 6 5 5 3 5 5 2 3 6 3 3 5 5 6 5
## [1111] 3 5 5 3 6 4 5 3 5 3 3 4 5 6 5 5 6 4 3 3 5 4 3 3 3 3 3 3 4 6 3 3 3 4 5 5 3
## [1148] 2 5 3 3 3 5 5 5 5 3 4 6 5 6 5 5 6 3 5 5 3 3 3 5 6 5 3 6 6 5 3 6 3 5 5 4 5
## [1185] 6 3 3 5 3 3 5 3 6 4 5 5 5 5 5 6 3 3 5 5 4 3 3 5 5 5 3 3 3 6 5 5 5 5 6 5 3
## [1222] 5 6 5 3 3 5 5 6 5 4 3 5 4 4 5 3 4 5 3 1 3 3 3 4 5 5 5 5 5 3 3 3 3 5 5 6 5
## [1259] 5 5 3 6 4 5 3 5 5 3 5 5 5 3 3 5 3 5 5 5 5 5 5 3 3 5 5 5 3 5 4 5 3 6 5 5 5
## [1296] 6 3 3 3 5 5 5 5 3 5 5 3 6 5 5 5 3 5 5 5 3 6 5 5 5 3 5 5 5 5 3 3 5 5 5 5 5
## [1333] 4 5 5 5 3 5 5 2 5 4 2 6 5 3 5 3 3 3 3 5 5 2 5 3 5 5 6 5 1 5 3 3 5 6 4 3 5
## [1370] 3 5 3 3 6 5 5 3 3 6 4 5 5 5 3 5 5 2 4 3 3 3 5 2 5 5 3 3 6 5 5 6 3 4 3 5 5
## [1407] 3 5 5 4 5 3 3 2 6 5 2 3 5 4 5 3 3 2 5 5 5 6 1 5 3 5 5 5 3 5 2 5 5 5 3 4 5
## [1444] 4 6 5 6 3 6 5 3 5 3 1 5 3 5 5 5 3 4 3 5 3 3 5 6 3 5 4 2 5 3 5 5 5 3 5 3 6
## [1481] 3 5 5 3 5 6 5 5 4 5 5 5 5 5 5 1 2 5 6 3 3 3 5 3 5 5 5 3 4 5 5 4 5 3 5 5 2
## [1518] 5 6 5 5 5 6 6 6 3 3 3 4 3 6 4 6 5 2 3 5 3 3 4 6 2 6 5 6 2 5 5 5 5 5 5 3 5
## [1555] 3 6 6 2 6 2 3 5 4 5 5 5 5 6 6 3 5 5 5 6 3 5 4 2 3 5 5 6 6 2 4 5 2 3 5 1 5
## [1592] 5 6 5 5 5 5 5 5 5 5 5 1 5 6 5 2 5 3 5 5 5 3 5 3 5 5 4 2 5 5 5 3 3 4 5 5 3
## [1629] 5 5 4 5 6 5 5 3 5 2 5 3 3 3 3 5 5 5 5 5 3 5 5 6 2 5 4 5 6 5 3 4 5 3 5 6 5
## [1666] 3 6 3 5 5 5 5 3 6 3 3 3 3 3 5 5 5 3 5 3 3 3 6 3 5 3 5 6 5 3 5 3 1 1 6 3 3
## [1703] 3 5 6 6 3 6 1 5 2 3 5 5 5 5 3 4 4 3 4 5 3 5 5 5 2 5 5 2 6 6 5 5 3 5 3 3 5
## [1740] 3 5 5 5 3 5 3 6 5 3 3 6 5 6 3 5 6 5 5 3 2 5 3 6 5 6 5 3 5 3 2 6 4 5 5 3 3
## [1777] 5 3 3 3 3 3 5 6 5 5 3 5 3 5 5 6 3 6 6 5 5 5 5 6 6 5 5 6 5 3 5 3 3 6 5 6 5
## [1814] 3 6 5 3 5 3 3 5 3 5 3 5 5 5 3 5 5 5 5 5 3 3 5 6 6 5 3 5 5 5 6 5 3 5 5 3 3
## [1851] 3 3 2 5 1 3 3 5 3 3 3 5 5 5 5 5 5 5 6 5 6 5 6 6 5 3 3 5 1 5 5 4 2 5 4 5 3
## [1888] 5 3 3 3 6 3 3 4 3 5 6 6 5 5 3 5 5 3 2 2 3 5 5 5 5 1 2 5 5 4 5 3 5 3 3 3 3
## [1925] 4 3 5 3 2 4 5 3 5 6 5 5 5 5 6 4 3 5 3 3 6 3 3 5 5 3 6 3 5 5 3 5 5 5 6 4 6
## [1962] 5 3 3 6 3 5 3 2 5 4 5 5 3 3 3 3 2 5 6 3 5 1 5 6 5 5 3 3 5 5 5 3 5 5 5 3 5
## [1999] 3 3
##
## Within cluster sum of squares by cluster:
## [1] 503852.1 323019.0 768792.8 323478.8 691495.1 344053.9
## (between_SS / total_SS = 82.1 %)
##
## Available components:
##
## [1] "cluster" "centers" "totss" "withinss" "tot.withinss"
## [6] "betweenss" "size" "iter" "ifault"
### Notes about the output of kmeans model
#### The output of the k-means model is a list
#### containing the cluster assignments, the cluster centers,
#### and other information.
#### The metric between_SS/total_SS helps evaluate how well the
#### clusters are separated. It represents the proportion of total
#### variance that is explained by the clusters,
#### which is 82.1% in our case here.
#STEP 7: Clustering Analysis
##First, Add cluster labels to the original dataframe
retail_segmentation <- retail_segmentation %>%
mutate(cluster = kmeans_model$cluster)
## View the first few rows with cluster assignment and see
## the new column of cluster in the data
head(retail_segmentation)
## Cust_No avg_order_size avg_order_freq crossbuy multichannel per_sale tenure
## 1 1 23.400000 2.2222222 3 2 0.00000000 3
## 2 2 34.260377 6.6250000 7 2 0.11111111 35
## 3 3 43.575641 4.8750000 5 2 0.07407407 12
## 4 4 26.316667 0.9000000 4 2 0.25000000 9
## 5 5 8.269231 1.0833333 3 1 0.50000000 40
## 6 6 21.500000 0.2222222 1 2 0.00000000 7
## return_rate married own_home household_size loyalty_card income age
## 1 0.1175214 1 1 1 1 35 47
## 2 0.2818684 1 1 3 1 140 70
## 3 0.2741769 1 0 4 0 35 21
## 4 0.1435508 0 0 1 1 35 62
## 5 0.0000000 0 0 2 0 140 21
## 6 0.0000000 0 1 1 1 80 21
## avg_mktg_cnt zip_code sales profit cluster
## 1 56.000000 21230 52.000000 -14.9600000 5
## 2 14.914286 22301 226.975000 106.8412857 2
## 3 20.083333 19002 212.431250 95.4017500 2
## 4 8.222222 22304 23.685000 6.1495333 5
## 5 1.350000 20124 8.958333 3.6458333 3
## 6 2.714286 22033 4.777778 0.4487302 5
## Scatter plot of clusters (using 'income' versus 'profit')
ggplot(retail_segmentation, aes(x = income, y = profit,
color = factor(cluster))) +
geom_point(size=3) +
labs(x = "Income", y = "Profit", color = "Cluster")
### Notes for interpretation: Cluster 1, 2, and 4 look more profitable
### across the levels of annual income.
##We can use the "kmeans_model$centers" to get the cluster centers
##that contains the average values of the variables for each cluster.
## See the average of the variables per cluster
kmeans_model$centers %>%
round(2) %>%
as.data.frame() %>%
tibble::rownames_to_column(var = "Cluster")
## Cluster tenure own_home married household_size income sales profit
## 1 1 17.82 0.50 0.43 2.79 96.07 433.91 204.78
## 2 2 15.16 0.61 0.45 3.29 60.48 205.77 89.99
## 3 3 16.18 0.52 0.53 2.76 127.77 22.37 4.39
## 4 4 18.75 0.40 0.62 2.57 125.35 111.75 44.18
## 5 5 11.61 0.62 0.40 2.97 45.79 17.76 0.52
## 6 6 15.29 0.55 0.48 2.77 46.05 85.81 32.06
## Notes on the results: Yes, Clusters 1, 2 and 4 are more profitable!!
## Enhanced cluster visualization using factoextra package
fviz_cluster(kmeans_model, data=cluster_data, geom="point",
ellipse.type="norm")
