How to create a multiple bar plot out of a m*n matrix

[punkz]

I would like to create a multiple bar plot out of a data frame that has an mn matrix structure. This will have a total of mn cells such that each cell will have a bar-plot of its equivalent value.

MN matrices are called incidence matrices as opposed to MM matrices which are called adjacency matrices.

I found out the way of plotting this for data frame with adjacency matrix (mm) matrix structure using igrpah package in R. However, I could not figure out vertices metadata for incidence (mn) matrix structure.
Could you please create a multiple bar plot for this type of data frame in R.
Provided below is an imaginary data frame of the number of tourists visiting 5 different countries for 12 months.

set.seed(19881003)
Australia <- (runif(12,0,1))
France <- (runif(12, 0,1) + Australia)
UnitedKingdom <- rbinom(12,2, 0.6)
USA <- rpois(12,0.06)+UnitedKingdom
Italy <- UnitedKingdom+USA+0.012
months <- c("Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug", "Sep", "Oct", "Nov", "Dec")
#
tourist.data <- data.frame(months, Australia, France, UnitedKingdom, USA, Italy)

I have also attached a sample chart. This chart differs from what I want in that the x-axis should represent 12 months instead of the countries

[FJCC]

Is this close enough?

library(ggplot2)
library(tidyr)
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, union
set.seed(19881003)
Australia <- (runif(12,0,1))
France <- (runif(12, 0,1) + Australia)
UnitedKingdom <- rbinom(12,2, 0.6)
USA <- rpois(12,0.06)+UnitedKingdom
Italy <- UnitedKingdom+USA+0.012
months <- c("Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug", "Sep", "Oct", "Nov", "Dec")
#
tourist.data <- data.frame(months, Australia, France, UnitedKingdom, USA, Italy)
  
tourist.data |> mutate(months=factor(months,levels = month.abb)) |> 
  pivot_longer(cols = -months) |>  
  ggplot(aes(months,value)) + geom_col() +
  facet_grid(name~months, scales = "free_x")

^{Created on 2024-02-28 with reprex v2.0.2}

[punkz]

Thank you. It is close enough. Could you think of it as a network plot or a heatmap plot? In case of adjacency matrix where the rows and columns are the same, I found out the way . Can you or anybody help me to plot country versus month plot in a way somewhat similar to the following country versus country plot:

set.seed(19881003)
Australia <- (runif(5,0,1))
France <- (runif(5, 0,1) + Australia)
UnitedKingdom <- rbinom(5,2, 0.6)
USA <- rpois(5,0.06)+UnitedKingdom
Italy <- UnitedKingdom+USA+0.012
country <- c("Australia", "France", "UnitedKingdom", "USA", "Italy")
touristflow <- data.frame(country, Australia, France, UnitedKingdom, USA, Italy)

library(tidyr)
main_data<- pivot_longer(touristflow,cols=c("Australia", "France", "UnitedKingdom", "USA","Italy"),
                   names_to = "dest_country", values_to = "tourist_flow (in millions)")
## Also prepare a dataframe of vertices and other attributes
support_data<- data.frame(region = 
                     c("Australia",
                       "France",
                       "UnitedKingdom",
                       "USA", "Italy"),
                 order1 = c(3,2,1,4,5)
                          )

## prepare for chart
svg_file <- "multiplebarplot.svg"
svg(svg_file, width=8, height=10)

library(igraph)
library(RColorBrewer)
## 
## prepare data
flowdat <- graph_from_data_frame(d=main_data, vertices = support_data,
                                 directed = TRUE)
flowdat_mat <- get.adjacency(flowdat, attr = "tourist_flow (in millions)", sparse = FALSE) 
maxmvalue <- max(flowdat_mat)
n <- nrow(flowdat_mat)
m <- n
##
par(mfrow = c(n,m), omi = c(1,4,4,2), mai=c(0,0,0,0))
color1 <- rgb(255,0,210, maxColorValue = 255)
color2 <- rgb(0,208,226, maxColorValue = 255)

## chart
for (i in 1:n) {
    for (j in 1:m) {
        plot (1:1, xlim= c(0,1), ylim = c(0,1), type = "n", axes = FALSE)
        if (i<j) mycolor <- color1
        if (i==j) mycolor <- "grey80"
        if (i>j) mycolor <- color2
        if(i==1) text(0.5, 1.2,support_data$region[j],
                      cex=2, xpd=NA, adj=0, srt=45,col=color1)
        if(j==1) text(-0.1, 0.5, support_data$region[i],
                      cex=2, xpd=NA, adj=1, col=color2)
        rect(0,0,1,1, col="grey95", border=NA)
        rect(0,0,1,flowdat_mat[i,j]/maxmvalue, col=mycolor,
             border=NA)
        text(0.5,0.5,format(round(flowdat_mat[i,j],2), nsmall=2), cex=1.5,
             col="grey40")
       
    }
}
mtext("Flow to:",side=3, outer = TRUE, cex=2.5, line=14, col=mycolor1, adj=0)
mtext("Tourist Flow (in millions)", side=3, outer = TRUE, cex=3, adj=1, at=0.4, line = 22, col="grey50", family="Lato Black")

dev.off()

[punkz]

Hi. Thank you for your help
I was just wondering if I could reorder the countries according to my choice. For instance I would like to place Italy first followed by France, Australia, USA and finally UnitedKingdom.
Any help is much appreciated

[FJCC]

You can set the order of the countries using the factor() function.

library(ggplot2)
library(tidyr)
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, union

set.seed(19881003)
Australia <- (runif(12,0,1))
France <- (runif(12, 0,1) + Australia)
UnitedKingdom <- rbinom(12,2, 0.6)
USA <- rpois(12,0.06)+UnitedKingdom
Italy <- UnitedKingdom+USA+0.012
months <- c("Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug", "Sep", "Oct", "Nov", "Dec")
#
tourist.data <- data.frame(months, Australia, France, UnitedKingdom, USA, Italy)

tourist.data |> mutate(months=factor(months,levels = month.abb)) |> 
  pivot_longer(cols = -months) |>
  mutate(name = factor(name, levels = c("Italy", "France", "Australia", "USA", "UnitedKingdom"))) |> 
  ggplot(aes(months,value)) + geom_col() +
  facet_grid(name~months, scales = "free_x")

^{Created on 2024-03-31 with reprex v2.0.2}