Twitter Pre-Module Assignment - Network Analytics
1. Objective and Data Description
For this assignment we will use Twitter data to explore network graphs. The dataset was collected in July 2014 from four major Twitter accounts (Snapchat, Dropbox and two others) as seed nodes, collecting their most recent 4000 followers. The same was done for each of these followers. Then, for each of these accounts, all the accounts they follow were collected. The data set includes 59,974 Twitter users (nodes) and 73,277 follower relationships (edges).
The data is available in the following files:
- graph complete.txt provides the edges of the graph in the form from => to. Each line is an edge where each node is separated by a space.
- graph complete.txt provides the edges of the graph in the form from => to. Each line is an edge where each node is separated by a space.
- graph subset.txt: a subset of the complete network. This file contains roughly 1% of the total number of edges (randomly selected). Each line is an edge where each node is separated by a space.
- graph subset.txt: a subset of the complete network. This file contains roughly 1% of the total number of edges (randomly selected). Each line is an edge where each node is separated by a space.
- ids to usernames.csv: maps the integer ids given in the two data files to the actual Twitter usernames of the users in our dataset. There are two comma separated fields in this file: the integer id and the string username.
2. Import Libraries
For this assignment, we will use the igraph package for network analysis.
pacman::p_load(igraph, DT, ggplot2, dplyr)3. Network Structure Visualization
Let us first plot the network by using the information in the file graph subset.txt. Note that this is not the complete network, but only a subset of its edges. By visualizing the graph, we can get an idea of the structure of the network we will be working on.
For better visualization, we will treat this data file as being in ncol format and turn on directed=TRUE. We also set vertex.size=1, vertex.color=“green”, layout=layout.kamada.kawai, vertex.label=NA, edge.arrow.size=.2, edge.curved =.1 and width=.5.
graph_subset <- read_graph("graph_subset.txt", format="ncol", directed = TRUE)
plot.igraph(graph_subset,
vertex.size = 1,
vertex.color = "green",
layout = layout.kamada.kawai,
vertex.label = NA,
edge.arrow.size = .4,
edge.curved = .1,
width = .5,
main = "Subset of Twitter Network (Kamada Kawai)")
Let us also try other graph layouts to see the network from different perspectives. There are many layouts available in the igraph package. We will go ahead to pick five alternative layouts: In Circle, On Sphere, Fruchterman-Reingold, Nicely and Tree.
layouts <- grep("^layout_", ls("package:igraph"), value=TRUE)[-1]
layouts## [1] "layout_as_bipartite" "layout_as_star" "layout_as_tree"
## [4] "layout_components" "layout_in_circle" "layout_nicely"
## [7] "layout_on_grid" "layout_on_sphere" "layout_randomly"
## [10] "layout_with_dh" "layout_with_drl" "layout_with_fr"
## [13] "layout_with_gem" "layout_with_graphopt" "layout_with_kk"
## [16] "layout_with_lgl" "layout_with_mds" "layout_with_sugiyama"plot.igraph(graph_subset,
vertex.size = 1,
vertex.color = "green",
layout = layout_in_circle,
vertex.label = NA,
edge.arrow.size = .4,
edge.curved = .1,
width = .5,
main = "Subset of Twitter Network (Layout In Circle)")
plot.igraph(graph_subset,
vertex.size = 1,
vertex.color = "green",
layout = layout_on_sphere,
vertex.label = NA,
edge.arrow.size = .4,
edge.curved = .1,
width = .5,
main = "Subset of Twitter Network (Layout on Sphere)")
plot.igraph(graph_subset,
vertex.size = 1,
vertex.color = "green",
layout = layout_with_fr,
vertex.label = NA,
edge.arrow.size = .4,
edge.curved = .1,
width = .5,
main = "Subset of Twitter Network (Fruchterman-Reingold)")
plot.igraph(graph_subset,
vertex.size = 1,
vertex.color = "green",
layout = layout_nicely,
vertex.label = NA,
edge.arrow.size = .4,
edge.curved = .1,
width = .5,
main = "Subset of Twitter Network (Layout Nicely)")
graph_subset1 <- read_graph("graph_subset.txt", format="ncol", directed = FALSE)
plot.igraph(graph_subset1,
vertex.size = 1,
vertex.color = "green",
layout = layout_as_tree(graph_subset1, circular = TRUE),
vertex.label = NA,
edge.arrow.size = .4,
edge.curved = .1,
width = .5,
main = "Subset of Twitter Network (Tree)")
We can make two interesting observations based on a visual inspection of the graphs. Firstly, the graph appears to be connected, i.e. there is a path to every node in the network. Secondly, there appears to be four key nodes - giant components that contains a significant fraction of all the nodes. This could be due to the fact that the data was collected from four major Twitter accounts.
Let us take a closer look at the graph. We begin by examining the user names. The table below shows the first few user names.
usernames <- read.csv("ids_to_usernames.csv", stringsAsFactors = FALSE)
head(usernames)## id name
## 1 0 Snapchat
## 2 1 insomniacevents
## 3 2 Dropbox
## 4 3 olympiacos_org
## 5 4 24_ida
## 6 5 IDDVxNext, we take a look at the relationships among the users.
V(graph_subset) # prints the vertices (i.e. Twitter users) ## + 1042/1042 vertices, named, from 2bc1c4e:
## [1] 70 0 186 219 477 530 543 545 636 746 931
## [12] 940 955 957 1115 1281 1355 1482 1753 1899 1957 2445
## [23] 2664 2794 3067 3389 3590 3710 3754 3771 3863 3896 3964
## [34] 3970 4050 1 4144 4232 4264 4362 4403 4404 4441 4473
## [45] 4497 4711 4807 4973 5098 5100 5108 5134 5169 5262 5303
## [56] 5396 5435 5439 5571 5616 5704 5714 5779 5826 5977 6034
## [67] 6168 6234 6256 6315 6327 6522 6621 7029 7049 7157 7221
## [78] 7341 7370 7497 7532 7787 7850 7995 8051 2 8098 8209
## [89] 8352 8359 8426 8539 8552 8562 8594 8758 8996 8998 9045
## [100] 9377 9411 9426 9471 9494 9506 9591 9816 9898 9971 10054
## + ... omitted several verticesE(graph_subset) # prints the edges (i.e. relationships)## + 734/734 edges from 2bc1c4e (vertex names):
## [1] 70 ->0 186 ->0 219 ->0 477 ->0 530 ->0 543 ->0 545 ->0 636 ->0
## [9] 746 ->0 931 ->0 940 ->0 955 ->0 957 ->0 1115->0 1281->0 1355->0
## [17] 1482->0 1753->0 1899->0 1957->0 2445->0 2664->0 2794->0 3067->0
## [25] 3389->0 3590->0 3710->0 3754->0 3771->0 3863->0 3896->0 3964->0
## [33] 3970->0 4050->1 4144->1 4232->1 4264->1 4362->1 4403->1 4404->1
## [41] 4441->1 4473->1 4497->1 4711->1 4807->1 4973->1 5098->1 5100->1
## [49] 5108->1 5134->1 5169->1 5262->1 5303->1 5396->1 5435->1 5439->1
## [57] 5571->1 5616->1 5704->1 5714->1 5779->1 5826->1 5977->1 6034->1
## [65] 6168->1 6234->1 6256->1 6315->1 6327->1 6522->1 6621->1 7029->1
## [73] 7049->1 7157->1 7221->1 7341->1 7370->1 7497->1 7532->1 7787->1
## + ... omitted several edgesdeg_in <- degree(graph_subset, mode="in") # Number of incoming relationships among users
deg_in_sorted <- sort.int(deg_in, decreasing=TRUE, index.return=FALSE)
head(deg_in_sorted)## 1 2 0 3 4406 7049
## 49 49 33 33 4 3(summary(deg_in))## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.0000 0.0000 0.0000 0.7044 1.0000 49.0000deg_out <- degree(graph_subset, mode="out") # Number of outgoing relationships among users
deg_out_sorted <- sort.int(deg_out, decreasing=TRUE, index.return=FALSE)
head(deg_out_sorted) ## 1 8599 5137 6738 1564 1288
## 13 3 3 3 3 3(summary(deg_out))## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.0000 0.0000 1.0000 0.7044 1.0000 13.0000ggplot() + geom_point(aes(c(1:length(deg_in)),deg_in)) +
ylab("Incoming Relationships") +
xlab("Users") +
ggtitle("Number of Incoming Relationships among Users") +
annotate("text", x = 100, y = 47, label = "insomniacevents (1)") +
annotate("text", x = 150, y = 51, label = "Dropbox (2)") +
annotate("text", x = 70, y = 30, label = "Snapchat (0)") +
annotate("text", x = 200, y = 35, label = "olympiacos_org (3)")
ggplot() + geom_point(aes(c(1:length(deg_out)),deg_out)) +
ylab("Outgoing Relationships") +
xlab("Users") +
ggtitle("Number of Outgoing Relationships among Users") +
annotate("text", x = 150, y = 12, label = "insomniacevents (1)") 
There are 1,042 vertices (i.e. Twitter users) and 734 edges (i.e. relationships) between them. We see that nodes 1 (insomniacevents), 2 (Dropbox), 0 (Snapchat) and 3 (olympiacos_org) stand out as having a significantly higher number of incoming relationships (49, 49, 33 and 33 respectively). The other nodes have 4 or less incoming relationships. Some nodes have no incoming relationship. The average number of incoming relationships per node is 0.7.
As for outgoing relationships, we see that node 1 (insomniacevents) stands out as having 13 outgoing relationships while the rest have 3 or less. Again, some nodes have no outgoing relationship. The average number of outgoing relationships per node is 0.7.
4. Data Analysis
We will now examine the complete graph contained in the file graph complete.txt and the username file ids_to_usernames.csv.
We want to plot the distribution of the number of followers of each user in our dataset (x-axis number of followers, y-axis number of nodes). For each edge, user a is said to be a follower of user b if there is some edge a => b.
- We start by counting the number of followers of each user using the table command. This is the same as the in-degree of each node in the graph. The table below shows the top users with the most followers. Again, we see nodes 0 (Snapchat), 1 (insomniacevents), 2 (Dropbox), and 3 (olympicos_org) stand out, with 4,020 followers each. The other nodes have 120 or less incoming relationships.
graph_complete <- read.table("graph_complete.txt", header = FALSE)
followers <- as.data.frame(table(graph_complete$V2))
names(followers) <- c("Node", "Number of Followers")
followers_usernames <- merge.data.frame(followers, usernames, by.x = 'Node', by.y = 'id')
followers_sorted <- followers_usernames %>%
arrange(desc(followers_usernames$`Number of Followers`))
head(followers_sorted)## Node Number of Followers name
## 1 0 4020 Snapchat
## 2 1 4020 insomniacevents
## 3 2 4020 Dropbox
## 4 3 4020 olympiacos_org
## 5 11705 120 elcapimar
## 6 1288 120 Kill_Joy7Of the 59,973 users, there are 27,457 with at least 1 relationship. This means that there are 32,516 (= 59973 - 27,457) users with 0 relationship. The average number of incoming relationships per node is 1.222.
graph_complete1 <- read_graph("graph_complete.txt", format="ncol", directed = TRUE)
deg_in_df <- data.frame(followers = degree(graph_complete1, mode="in"), userid = names(degree(graph_complete1)))
deg_in_df_sorted <- sort.int(deg_in_df$followers, decreasing=TRUE, index.return=FALSE)
table(deg_in_df_sorted)## deg_in_df_sorted
## 0 1 2 3 4 5 6 7 8 9 10 11
## 32516 24454 1368 434 249 142 83 65 57 39 27 30
## 12 13 14 15 16 17 18 19 20 21 22 23
## 21 21 12 16 14 10 8 10 18 4 1 7
## 24 25 26 27 28 29 30 31 32 33 34 35
## 7 6 3 4 4 4 1 5 3 6 3 5
## 36 37 38 39 40 41 42 43 45 46 47 48
## 3 3 4 3 85 3 1 3 1 1 4 1
## 49 50 51 52 53 55 56 57 58 59 60 61
## 4 1 4 1 1 1 2 1 3 1 55 1
## 63 65 67 68 70 73 75 77 79 80 86 96
## 2 1 1 1 2 1 3 2 2 37 1 1
## 97 99 100 102 108 111 116 120 4020
## 1 2 32 1 1 2 1 31 4summary(deg_in_df_sorted)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.000 0.000 0.000 1.222 1.000 4020.000We can validate this by examining the in-degree of each node.
V(graph_complete1) ## + 59973/59973 vertices, named, from 310da36:
## [1] 4 0 5 6 7 8 9 10 11 12 13
## [12] 14 15 16 17 18 19 20 21 22 23 24
## [23] 25 26 27 28 29 30 31 32 33 34 35
## [34] 36 37 38 39 40 41 42 43 44 45 46
## [45] 47 48 49 50 51 52 53 54 55 56 57
## [56] 58 59 60 61 62 63 64 65 66 67 68
## [67] 69 70 71 72 73 74 75 76 77 78 79
## [78] 80 81 82 83 84 85 86 87 88 89 90
## [89] 91 92 93 94 95 96 97 98 99 100 101
## [100] 102 103 104 105 106 107 108 109 110 111 112
## + ... omitted several verticesE(graph_complete1) ## + 73277/73277 edges from 310da36 (vertex names):
## [1] 4 ->0 5 ->0 6 ->0 7 ->0 8 ->0 9 ->0 10 ->0 11 ->0 12 ->0 13 ->0
## [11] 14 ->0 15 ->0 16 ->0 17 ->0 18 ->0 19 ->0 20 ->0 21 ->0 22 ->0 23 ->0
## [21] 24 ->0 25 ->0 26 ->0 27 ->0 28 ->0 29 ->0 30 ->0 31 ->0 32 ->0 33 ->0
## [31] 34 ->0 35 ->0 36 ->0 37 ->0 38 ->0 39 ->0 40 ->0 41 ->0 42 ->0 43 ->0
## [41] 44 ->0 45 ->0 46 ->0 47 ->0 48 ->0 49 ->0 50 ->0 51 ->0 52 ->0 53 ->0
## [51] 54 ->0 55 ->0 56 ->0 57 ->0 58 ->0 59 ->0 60 ->0 61 ->0 62 ->0 63 ->0
## [61] 64 ->0 65 ->0 66 ->0 67 ->0 68 ->0 69 ->0 70 ->0 71 ->0 72 ->0 73 ->0
## [71] 74 ->0 75 ->0 76 ->0 77 ->0 78 ->0 79 ->0 80 ->0 81 ->0 82 ->0 83 ->0
## [81] 84 ->0 85 ->0 86 ->0 87 ->0 88 ->0 89 ->0 90 ->0 91 ->0 92 ->0 93 ->0
## [91] 94 ->0 95 ->0 96 ->0 97 ->0 98 ->0 99 ->0 100->0 101->0 102->0 103->0
## + ... omitted several edgesdeg_in1 <- degree(graph_complete1, mode="in")
deg_in1_sorted <- sort.int(deg_in1, decreasing=TRUE, index.return=FALSE)
head(deg_in1_sorted)## 0 1 2 3 266 336
## 4020 4020 4020 4020 120 120ggplot() + geom_point(aes(c(1:length(deg_in1)),deg_in1)) +
ylab("Incoming Relationships") +
xlab("Users") +
ggtitle("Number of Incoming Relationships among Users") +
annotate("text", x = 20000, y = 3500, label = 'Nodes 0-3 stand out with 4,020 followers each.')
- Next, we apply the same process to count the number of users (nodes) that have a particular number of followers, that is, the number of outgoing relationships that have a destination node in the graph.
nodes <- as.data.frame(table(graph_complete$V1))
names(nodes) <- c("Node", "Number of Users")
nodes_sorted <- nodes %>%
arrange(desc(nodes$`Number of Users`))
head(nodes_sorted)## Node Number of Users
## 1 1 1976
## 2 173 121
## 3 655 121
## 4 1212 121
## 5 1288 121
## 6 1804 121We see that node 1 (insomniacevents) stands out as having 1,976 outgoing relationships while the rest have 121 or less. Of the 59,973 users, there are 37,679 with at least 1 outgoing relationship. This means that there are 22,294 (= 59973 - 37,679) users with 0 relationship. The average number of outgoing relationships per node is 1.222.
deg_out_df <- data.frame(followers = degree(graph_complete1, mode="out"), userid = names(degree(graph_complete1)))
deg_out_df_sorted <- sort.int(deg_out_df$followers, decreasing=TRUE, index.return=FALSE)
table(deg_out_df_sorted)## deg_out_df_sorted
## 0 1 2 3 4 5 6 7 8 9 10 11
## 22294 36155 725 146 53 29 18 12 6 4 8 12
## 12 13 14 15 16 17 18 19 20 21 23 24
## 9 4 5 5 4 5 1 2 3 22 1 2
## 25 26 27 28 29 30 32 33 34 36 37 38
## 4 2 1 1 2 1 1 2 1 1 4 1
## 40 41 42 44 45 46 47 48 49 50 51 52
## 2 123 2 3 1 2 1 5 1 1 3 2
## 53 54 55 56 59 60 61 62 63 67 70 72
## 2 2 2 1 2 2 68 2 1 1 3 1
## 73 77 78 80 81 84 89 90 95 98 99 100
## 1 1 1 2 59 1 1 2 1 3 2 1
## 101 104 105 113 121 1976
## 59 1 2 2 54 1summary(deg_out_df_sorted)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.000 0.000 1.000 1.222 1.000 1976.000And we can validate this by examining the out-degree of each node.
followers_out <- data.frame(followers = degree(graph_complete1, mode = "out"),
userid = names(degree(graph_complete1)))
followers_out_usernames <- merge.data.frame(followers_out, usernames, by.x = 'userid', by.y = 'id')
followers_out_usernames_sorted <- followers_out_usernames %>%
arrange(desc(followers_out_usernames$followers))
head(followers_out_usernames_sorted)## userid followers name
## 1 1 1976 insomniacevents
## 2 10098 121 RachelNIntuit
## 3 10181 121 TnahaSingh
## 4 10193 121 tuprasanti
## 5 10350 121 xwx28
## 6 10856 121 AutotaskLendeg_out1 <- degree(graph_complete1, mode="out")
deg_out1_sorted <- sort.int(deg_out1, decreasing=TRUE, index.return=FALSE)
head(deg_out1_sorted)## 1 173 655 1212 1288 1804
## 1976 121 121 121 121 121ggplot() + geom_point(aes(c(1:length(deg_out1)),deg_out1)) +
ylab("Outgoing Relationships") +
xlab("Users") +
ggtitle("Number of Outgoing Relationships among Users") +
annotate("text", x = 18000, y = 1800, label = 'Node 1 stand out with 1,976 outgoing relationships.')
Combining the two, we can see that node 1 has the most number of total relationships (5,996). Dropbox and Olympiacos_org, and Snapchat have the next three highest number of relationships. The rest are have 241 or less relationships.
followers_total <- data.frame(followers = degree(graph_complete1, mode = "total"),
userid = names(degree(graph_complete1)))
followers_total_usernames <- merge.data.frame(followers_total, usernames, by.x = 'userid', by.y = 'id')
followers_total_usernames_sorted <- followers_total_usernames %>%
arrange(desc(followers_total_usernames$followers))
head(followers_total_usernames_sorted)## userid followers name
## 1 1 5996 insomniacevents
## 2 2 4125 Dropbox
## 3 3 4073 olympiacos_org
## 4 0 4071 Snapchat
## 5 11705 241 elcapimar
## 6 1288 241 Kill_Joy7- We can now go ahead to plot the distribution. We want to have on the x-axis the number of followers, and on y-axis how many users have that number of followers. This can be plotted using the geom_density function in ggplot. There are 32,516 users with 0 relationship, 24,454 users with 1 follower and 4 users have 4,020 followers.
ggplot(deg_in_df, aes(x=followers)) +
geom_density(fill="seagreen3") +
labs(title="Density Distribution of Followers",
x = "Number of Followers",
y="Density") 
count <- data.frame(table((followers$`Number of Followers`)))
count## Var1 Freq
## 1 1 24454
## 2 2 1368
## 3 3 434
## 4 4 249
## 5 5 142
## 6 6 83
## 7 7 65
## 8 8 57
## 9 9 39
## 10 10 27
## 11 11 30
## 12 12 21
## 13 13 21
## 14 14 12
## 15 15 16
## 16 16 14
## 17 17 10
## 18 18 8
## 19 19 10
## 20 20 18
## 21 21 4
## 22 22 1
## 23 23 7
## 24 24 7
## 25 25 6
## 26 26 3
## 27 27 4
## 28 28 4
## 29 29 4
## 30 30 1
## 31 31 5
## 32 32 3
## 33 33 6
## 34 34 3
## 35 35 5
## 36 36 3
## 37 37 3
## 38 38 4
## 39 39 3
## 40 40 85
## 41 41 3
## 42 42 1
## 43 43 3
## 44 45 1
## 45 46 1
## 46 47 4
## 47 48 1
## 48 49 4
## 49 50 1
## 50 51 4
## 51 52 1
## 52 53 1
## 53 55 1
## 54 56 2
## 55 57 1
## 56 58 3
## 57 59 1
## 58 60 55
## 59 61 1
## 60 63 2
## 61 65 1
## 62 67 1
## 63 68 1
## 64 70 2
## 65 73 1
## 66 75 3
## 67 77 2
## 68 79 2
## 69 80 37
## 70 86 1
## 71 96 1
## 72 97 1
## 73 99 2
## 74 100 32
## 75 102 1
## 76 108 1
## 77 111 2
## 78 116 1
## 79 120 31
## 80 4020 4Let’s see the distribution more clearly by dropping the outliers.
followers_outliers <- deg_in_df[(deg_in_df$followers<4020),] # remove outliers
ggplot(followers_outliers, aes(x=followers)) +
geom_density(fill="seagreen3") +
labs(title="Density Distribution of Followers (without outliers)",
x = "Number of Followers",
y="Density") 
- Let’s transform the x-axis of the previous graph to logscale, to get a better understanding of the distribution. Note here some users that have 0 followers. This means that using the log of the x-axis will fail since log(0) will not be valid. Hence, we replace 0 with 0.1.
deg_in_df[(deg_in_df$followers ==0),'followers'] <- 0.1
ggplot(deg_in_df, aes(x=log(followers))) +
geom_density(fill="seagreen3") +
labs(title="Log-Density Distribution of Followers",
x = "Number of Followers",
y="Log-Density") 
Let’s also see the log-density distribution more clearly by dropping the outliers.
followers_outliers[(followers_outliers$followers == 0),'followers'] <- 0.1
ggplot(followers_outliers, aes(x=log(followers))) +
geom_density(fill="seagreen3") +
labs(title="Log-Density Distribution of Followers (without outliers)",
x = "Number of Followers",
y="Log-Density") 
The distribution is highly skewed to the right. Most users have very few followers but there are a few users with many followers. Ths is commonly seen in social networks. This skewness is seen by the low mean but high standard deviation of followers (see part 3 below).
- The average number of followers is 1.276 per user and the standard deviation is 33.25. The large disperson around the mean shows the skewness of the distribution.
round(mean(deg_in_df$followers),3)## [1] 1.276round(sd(deg_in_df$followers),2)## [1] 33.25- Finally, we report the Twitter usernames of the top 10 users with the most followers. We print users with multiple ties.
followers_top10 <- followers_sorted[followers_sorted$`Number of Followers` %in%
unique(followers_sorted$`Number of Followers`)[1:10],]
followers_top10## Node Number of Followers name
## 1 0 4020 Snapchat
## 2 1 4020 insomniacevents
## 3 2 4020 Dropbox
## 4 3 4020 olympiacos_org
## 5 11705 120 elcapimar
## 6 1288 120 Kill_Joy7
## 7 14978 120 JohnIrons95
## 8 1804 120 sydneyhbrodsky
## 9 2337 120 Can1ffs_bae
## 10 2370 120 mcclainxkylie
## 11 266 120 091Jacko
## 12 2796 120 redBONEmango_YA
## 13 336 120 byers_alec
## 14 3419 120 glasstablegirlz
## 15 3883 120 hannabaldovino
## 16 3916 120 jillian_1015
## 17 4098 120 kjrasing
## 18 4139 120 SongsThisWeek
## 19 4146 120 AminGa123
## 20 4164 120 fatimeDuraj
## 21 4293 120 MarvinaPete
## 22 4959 120 djshawnjay
## 23 5183 120 Massy_DeeJay
## 24 6313 120 Nique_Kee
## 25 6588 120 Enzolibo
## 26 6716 120 JayFulk
## 27 6984 120 miklit23
## 28 7300 120 tha__symbolic
## 29 7343 120 __rossgeller
## 30 7382 120 jesse_jamal
## 31 7525 120 BOXFEDmusic
## 32 7686 120 deephousCo
## 33 7970 120 GOvOSE
## 34 8673 120 PetjaSairanen
## 35 9513 120 Asit_Tewari
## 36 11163 111 namfow
## 37 3355 108 sammeyer98
## 38 7448 102 djMikeHawkins
## 39 1100 100 adamzfurniture
## 40 11298 100 cocaine345
## 41 11444 100 Steenberger
## 42 11686 100 Brendon_Smale
## 43 1212 100 brady1995
## 44 13443 100 ArgirisKarvelas
## 45 13959 100 ComEstadios
## 46 14141 100 Tommac1602
## 47 173 100 LuArDanni
## 48 1940 100 MaricelaSGarcia
## 49 220 100 gabider
## 50 2432 100 thetli8
## 51 2881 100 jodiieAnnex
## 52 3343 100 haley_harloff
## 53 3958 100 hesperfection
## 54 4406 100 CTG757
## 55 4730 100 Baharum135LC
## 56 5265 100 MikeBeEasy
## 57 5634 100 15serranista
## 58 5687 100 _Lyss18
## 59 5758 100 Goombaaabeth
## 60 6459 100 jeremysk82
## 61 6491 100 Callie_Mccleod
## 62 6665 100 AlanSalazarx_O
## 63 7049 100 paulinaxo_
## 64 7583 100 17Sihota
## 65 7878 100 bunnytheraver
## 66 8162 100 Natashamarko
## 67 9040 100 cristinaportalu
## 68 9141 100 HTMLCOIN
## 69 9266 100 marialu16955736
## 70 9812 100 pbower10
## 71 2581 99 David_Oz5
## 72 2675 99 jujuchaudin7
## 73 4036 97 Sean_Chell
## 74 7397 96 istekovic
## 75 6925 86 HardwellOnMe