#load data
setwd("~/Desktop/Network Analytics")
network_data_subset <-read.table(file = "graph_subset_rank1000.txt")
names(network_data_subset) <- c("from", "to")
#convert data
network_data_subset <-
graph_from_data_frame(network_data_subset, directed = FALSE, vertices = NULL)
#Plotting the network
plot.igraph(
network_data_subset,
layout = layout.kamada.kawai,
vertex.color = 'blue',
vertex.label = NA,
vertex.size = 4,
edge.curved = .1,
asp = -1,
margin = -.03
)
title("Network Data Plot",
cex.main = .75,
col.main = "Black")
#commments - there seems to be a large gap in the bottom left quadrant of the network data plot. The network data plot also seems to be clustered towards the edges
#using subset to plot top 1000
#Load data and convert
network_data_subset_cc <-read.table(file = "graph_subset_rank1000_CC.txt")
names(network_data_subset_cc) <- c("from", "to")
network_data_subset_cc <-
graph_from_data_frame(network_data_subset_cc, directed = F, vertices = NULL)
plot.igraph(
network_data_subset_cc,
layout = layout.kamada.kawai,
edge.width = 2,
vertex.color = V(network_data_subset_cc),
vertex.label = NA,
vertex.size = 3,
vertex.label.cex = .5,
edge.curved = .1,
asp = -1,
margin = -.03
)
title("Largest connected component Components of Top-Ranked Products",
cex.main = .75,
col.main = "Black")
#4 Data analysis
net_all <- read.table(file = "graph_complete.txt")
names(net_all) <- c("from", "to")
data_all_from<-as.data.frame(net_all$from)
data_all_to<-as.data.frame(net_all$to)
# Calculate Out & In Degrees
# convert data to grapnh object
net_all <-graph_from_data_frame(net_all, directed = TRUE, vertices = NULL)
# Calculate out degree using degree command
net_all$degree.out <-
degree(
net_all,
mode = "out"
)
# Calculate In degree using degree command
net_all$degree.in <-
degree(
net_all,
mode = "IN"
)
#Plot the out-degree distribution
# create data frame
net_outdeg<-as.data.frame(net_all$degree.out)
net_outdeg<-cbind.data.frame(rownames(net_outdeg),net_outdeg)
names(net_outdeg) <- c("from", "out degree")
mytheme <- theme_bw() +
theme(
text = element_text(family = "Times New Roman"),
panel.border = element_blank(),
panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
axis.line = element_line(color = "grey"),
axis.text.x = element_text(size = 10, angle = 0),
axis.text.y = element_text(size = 10, angle = 0),
axis.title.y = element_text(size = 10),
axis.title.x = element_text(size = 10)
)
# Plot distribution
ggplot(net_outdeg, aes(x = net_outdeg$`out degree`)) +
geom_histogram(
stat = "bin" ,
binwidth = 1,
aes(fill = ..count..),
show.legend = FALSE
) +
mytheme +
labs(title = "Out-Degree Distribution",
x = "Number of Similar Products",
y = "Number of Nodes") +
scale_fill_gradient(low = "grey", high = "black")
# convert out degree to data frame
net_indeg<-as.data.frame(net_all$degree.in)
# add rownames (i.e. From)
net_indeg<-cbind.data.frame(rownames(net_indeg),net_indeg)
names(net_indeg) <- c("from", "in degree")
ggplot(net_indeg, aes(x = net_indeg$`in degree`)) +
geom_histogram(
stat = "bin",
binwidth = 1,
aes(fill = ..count..),
show.legend = FALSE
) +
mytheme +
scale_y_continuous(labels = scales::comma) +
labs(title = "In-Degree Distribution (similar product limit = 30)",
x = "Number of Similar Products",
y = "Number of Nodes") +
scale_fill_gradient(low = "grey", high = "black") +
xlim (-1, 30)
#observation - the new distribution is right skewed, compared the the previous distribution which was left skewed
ggplot(net_indeg, aes(x = net_indeg$`in degree`)) +
geom_histogram(
stat = "bin",
binwidth = 1,
aes(fill = ..count..),
show.legend = FALSE
) +
mytheme +
scale_y_continuous(labels = scales::comma) +
labs(title = "In-Degree Distribution",
x = "Number of Similar Products",
y = "Number of Nodes") +
scale_fill_gradient(low = "grey", high = "black")
#Transform the x-axis of the previous graph to log scale, to get a better understanding of the distribution. Note here that you should have some products with 0 inbound links. This means that using the log of the x-axis will fail since log(0) will not be valid. Due to this, you should replace 0 with 0.1. Comment on what you observe.
# Replace 0 with 0.1
net_indeg_log <- as.data.frame(net_indeg)
net_indeg_log[net_indeg_log == 0] <- 0.1
# plot graph
ggplot(net_indeg_log, aes(x = net_indeg_log$`in degree`)) +
geom_histogram(
stat = "bin",
binwidth = 1,
aes(fill = ..count..),
show.legend = FALSE
) +
mytheme +
scale_x_log10() +
scale_y_continuous(labels = scales::comma) +
labs(title = "In-Degree Distribution Log Transform of similar products",
x = "Log (Number of Similar Products)",
y = "Number of Nodes") +
scale_fill_gradient(low = "grey", high = "black")
#comment on what you observe - after log transforming the distribtuion looks very different, it is not heavily right skewed compared to the previous distribution
#Compute the average number of inbound co-purchase links, the standard deviation, and the maximum. Comment on the result.
paste("The Average number of inbound co-purchase links is:",
round(mean(net_indeg$`in degree`), 1))
## [1] "The Average number of inbound co-purchase links is: 3.4"
paste("The Standard deviation of inbound co-purchase links is:",
round(sd(net_indeg$`in degree`), 1))
## [1] "The Standard deviation of inbound co-purchase links is: 6"
paste("The maximum of inbound co-purchase links is:", max(net_indeg$`in degree`))
## [1] "The maximum of inbound co-purchase links is: 549"
#Report the names of the 10 products with the most inbound co-purchase links.
# read product names as a table from.txt file
data_title <- read.csv(file = "id_to_titles.txt", sep = "")
# rename columns
names(data_title) = c("product", "name")
# join in-degree data frame with Title
df_top10_full <- net_indeg %>%
inner_join(data_title, by = c("from" = "product"))
# Create Data frame with top 10
df_top10 <-
as.data.frame(head(df_top10_full[order(df_top10_full$`in degree`, decreasing = T),],n=10))
#Report the names of the 10 products with the most inbound co-purchase links.
paste(df_top10$name)
## [1] "Laura"
## [2] "Diagnostic and Statistical Manual of Mental Disorders DSM-IV-TR (Text Revision) (Diagnostic and Statistical Manual of Mental Disorders)"
## [3] "Publication Manual of the American Psychological Association, Fifth Edition"
## [4] "The Great Gatsby"
## [5] "1001 Most Useful Spanish Words (Beginners' Guides)"
## [6] "It Works"
## [7] "Brown Bear, Brown Bear, What Do You See?"
## [8] "Easy Spanish Phrase Book: Over 770 Basic Phrases for Everyday Use"
## [9] "The Prince"
## [10] "The TEMPEST"