Create a chart from different data

Question

I need help to create a chart. I explain better.

I created 10 random graphs, each with N nodes. I have done that for N = 10^3, 10^4, 10^5. So in total 30 graphs.

To each of them I found the percentage of multilinks and selfloops they have.

Now I would like to create a single graph that shows the percentage in function of the number of nodes. So something like:

So I have a 3 lists: - listNets containing 30 graphs - listSelf containing the percentage of selfloops - listMul containing the percentage of multilinks

This is what I did:

listN <- c((10^3), (10^4), (10^5))

# list of networks
listNets <- vector(mode = "list", length = 0) 
# list of percentage of selfloops
listSelf <- vector(mode = "list", length = 0)
#list of percentage of multilinks
listMul <- vector(mode = "list", length = 0)

...

for(N in listN) {

    ...

    net <- graph_from_adjacency_matrix(adjmatrix = adjacency_matrix, mode = "undirected") # it's work, infact if I plot it i saw a correct networks 
    listNets <- c(listNets, net) # I add net to list of networks
    x11()
    plot(net, layout = layout.circle(net))

    ...

    # I find self-loops e multilinks
    netmatr <- as_adjacency_matrix(net, sparse = FALSE)
    num_selfloops <- sum(diag(netmatr))
    num_multilinks <- sum(netmatr > 1)

    # I find percentage
    per_self <- ((num_selfloops/num_vertices)*100)
    per_mul <- ((num_multilinks/num_edges)*100)

    listSelf <- c(listSelf, per_self) 
    listMul <- c(listMul, per_mul)
}

Now if I print listNets in this way I have something strange:

> print(listNets)
[[1]]
[1] 9

[[2]]
[1] FALSE

[[3]]
[1] 7 6 3 8 8 8

[[4]]
[1] 0 1 2 4 5 7

[[5]]
[1] 2 1 0 3 4 5

[[6]]
[1] 0 1 2 3 4 5

[[7]]
 [1] 0 0 0 0 1 1 1 2 3 6

[[8]]
 [1] 0 1 2 3 3 4 5 5 6 6

[[9]]
[[9]][[1]]
[1] 1 0 1

[[9]][[2]]
named list()

[[9]][[3]]
list()

[[9]][[4]]
list()


[[10]]
<environment: 0x000000001a6284a8>

[[11]]
[1] 9

[[12]]
[1] FALSE

[[13]]
[1] 2 5 8 8 7 8

[[14]]
[1] 0 1 3 4 6 7

[[15]]
[1] 0 1 4 2 3 5

[[16]]
[1] 0 1 2 3 4 5

[[17]]
 [1] 0 0 0 1 1 1 2 2 3 6

[[18]]
 [1] 0 1 2 2 3 4 4 5 6 6

[[19]]
[[19]][[1]]
[1] 1 0 1

[[19]][[2]]
named list()

[[19]][[3]]
list()

[[19]][[4]]
list()


[[20]]
<environment: 0x000000001a859e28>

...

Instead if I print the other two lists (listSelf and listMult everything is ok).

Now, how can I plot this data?

I read about dataframes, but I don't understand how to use it in my case. Can someone help me please?

I tried to bring me back by writing a possible result table on a csv file by hand and tried to plot it to see if I was going in the right direction.

This is the code and this the result. Note: The table I created by hand and I invented the percentages.

> df <- read.csv("./table.csv", sep = ",")  # read csv file 
> df
      N perSelf perMul
1  10^3       2      1
2  10^3       5      1
3  10^3      98     15
4  10^3      50     51
5  10^3      41     52
6  10^3      21    100
7  10^3      36     80
8  10^3      70     20
9  10^3      80     55
10 10^3     100     44
11 10^4       2      1
12 10^4       5     18
13 10^4     100     20
14 10^4      50     51
15 10^4      51     52
16 10^4      21    100
17 10^4      36     80
18 10^4      70     20
19 10^4      73     85
20 10^4     100     98
21 10^5     100     10
22 10^5       5      1
23 10^5      98     15
24 10^5      50     51
25 10^5      41     52
26 10^5      21     85
27 10^5      36     80
28 10^5      65     20
29 10^5      80     55
30 10^5     100     44

There is something wrong.

Thanks a lot

---

The code is:

# create a matrix from a list (list_all)
mat <- matrix(unlist(list_all), 
              unique(lengths(list_all)),
              dimnames = list(NULL, c("N", "% selfloops", "% multilinks")))

# convert matrix to data frame
df <- as.data.frame(x = mat, row.names = NULL) 
df

# plot
dflong <- melt(df, id.vars = 'N')

x11()
ggplot(dflong, aes(x = N, y = value, color = variable)) +
  geom_point(size = 5, alpha = 0.7, position = position_dodge(width = 0.3)) +
  scale_x_discrete(labels = parse(text = as.character(unique(dflong$N)))) +
  scale_y_continuous('', breaks = seq(0, 100, 25), labels = paste(seq(0, 100, 25), '%')) +
  scale_color_manual('', values = c('red', 'blue'),
                     labels = c('Percentage of selfloop','Percentage of multilinks')) +
  theme_minimal(base_size = 14)

df is:

   N % selfloops % multilinks
1 10   11.111111      0.00000
2 10   11.111111      0.00000
3 10    0.000000      0.00000
4 20    0.000000      0.00000
5 20    0.000000     15.38462
6 20    0.000000      0.00000
7 30    3.448276      0.00000
8 30    3.448276      0.00000
9 30    0.000000      0.00000

Answer 1

Taking your df dataframe as a starting point, you can get the desired result in two steps:

1) Reshape your data into long format with reshape2:

library(reshape2)
dflong <- melt(df, id.vars = 'N')

2) Plot the data with ggplot2:

library(ggplot2)
ggplot(dflong, aes(x = N, y = value, color = variable)) +
  geom_point(size = 5, alpha = 0.7, position = position_dodge(width = 0.3)) +
  scale_x_discrete(labels = parse(text = as.character(unique(dflong$N)))) +
  scale_y_continuous('', breaks = seq(0,100,25), labels = paste(seq(0,100,25),'%')) +
  scale_color_manual('', values = c('red','blue'), 
                     labels = c('Percentage of selfloop','Percentage of multilinks')) +
  theme_minimal(base_size = 14)

which gives:

I used a transparency (alpha = 0.7) in order to be able to see where points overlap.

---

In response to your comment and the second example in the question:

You have to alter the ggplot2 code a bit:

• Change the x variable in the aes to a factor.
• There is no need to parse the text for the labels anymore, thus that part can be removed.
• Adjust the values and breaks in the y-scale.

The following code:

ggplot(dflong, aes(x = factor(N), y = value, color = variable)) +
  geom_point(size = 5, alpha = 0.5, position = position_dodge(width = 0.3)) +
  xlab('N') +
  scale_y_continuous('', breaks = seq(0, 20, 5), 
                     labels = paste(seq(0, 20, 5), '%'),
                     limits = c(0,20)) +
  scale_color_manual('', 
                     values = c('red', 'blue'),
                     labels = c('Percentage of selfloop','Percentage of multilinks')) +
  theme_minimal(base_size = 14)

will give you:

---

Used data:

df <- structure(list(N = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L), .Label = c("10^3", "10^4", "10^5"), class = "factor"), 
                     perSelf = c(2L, 5L, 98L, 50L, 41L, 21L, 36L, 70L, 80L, 100L, 2L, 5L, 100L, 50L, 51L, 21L, 36L, 70L, 73L, 100L, 100L, 5L, 98L, 50L, 41L, 21L, 36L, 65L, 80L, 100L), 
                     perMul = c(1L, 1L, 15L, 51L, 52L, 100L, 80L, 20L, 55L, 44L, 1L, 18L, 20L, 51L, 52L, 100L, 80L, 20L, 85L, 98L, 10L, 1L, 15L, 51L, 52L, 85L, 80L, 20L, 55L, 44L)), 
                .Names = c("N", "perSelf", "perMul"), class = "data.frame", row.names = c(NA, -30L))