-
- -
-
Researchers network and migration flows
-
- A few data analytics ideas from Data-to-Viz.com
-
- -
-
- 
-
- 
-
- 
-
- 
-
-
-
-
-
-
-
- Researchers network and migration flows
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
+ Researchers network and migration flows
+
+
+
+
+
+
+
+
+
+ Adjacency and incidence matrix provide relationship between several nodes. The information they contain can have different nature, thus this document will consider several examples:
+Adjacency and incidence matrices provide relationship between several +nodes. The information they contain can have different nature, thus this +document will consider several examples:
directed and weighted. Like the number of people migrating from one country to another. Data used comes from this scientific publication from Gui J. Abel.directed and
+weighted. Like the number of people migrating from one
+country to another. Data used comes from this scientific
+publication from Gui J. Abel.# Libraries
-library(tidyverse)
-library(hrbrthemes)
-library(circlize)
-library(kableExtra)
-options(knitr.table.format = "html")
-library(viridis)
-library(igraph)
-library(ggraph)
-library(colormap)
-
-# Load dataset from github
-data <- read.table("https://raw.githubusercontent.com/holtzy/data_to_viz/master/Example_dataset/13_AdjacencyDirectedWeighted.csv", header=TRUE)
-
-# show data
-data %>% head(3) %>% select(1:3) %>% kable() %>%
- kable_styling(bootstrap_options = "striped", full_width = F)# Libraries
+library(tidyverse)
+library(hrbrthemes)
+library(circlize)
+library(kableExtra)
+options(knitr.table.format = "html")
+library(viridis)
+library(igraph)
+library(ggraph)
+library(colormap)
+
+# Load dataset from github
+data <- read.table("https://raw.githubusercontent.com/holtzy/data_to_viz/master/Example_dataset/13_AdjacencyDirectedWeighted.csv", header=TRUE)
+
+# show data
+data %>% head(3) %>% select(1:3) %>% kable() %>%
+ kable_styling(bootstrap_options = "striped", full_width = F)undirected and unweighted. I will consider all the co-authors of a researcher and study who is connected through a common publication. Data have been retrieved using the scholar package, the pipeline is describe in this github repository. The result is an adjacency matrix with about 100 researchers, filled with 1 if they have published a paper together, 0 otherwise.undirected and
+unweighted. I will consider all the co-authors of a
+researcher and study who is connected through a common publication. Data
+have been retrieved using the scholar package, the
+pipeline is describe in this github
+repository. The result is an adjacency matrix with about 100
+researchers, filled with 1 if they have published a paper together, 0
+otherwise.# Load data
-#dataUU <- read.table("https://raw.githubusercontent.com/holtzy/data_to_viz/master/Example_dataset/13_AdjacencyUndirectedUnweighted.csv", header=TRUE)
-dataUU <- read.table("../Example_dataset/13_AdjacencyUndirectedUnweighted.csv", header=TRUE)
-
-# show data
-dataUU %>% head(3) %>% select(1:4) %>% kable() %>%
- kable_styling(bootstrap_options = "striped", full_width = F)# Load data
+dataUU <- read.table("https://raw.githubusercontent.com/holtzy/data_to_viz/master/Example_dataset/13_AdjacencyUndirectedUnweighted.csv", header=TRUE)
+
+# show data
+dataUU %>% head(3) %>% select(1:4) %>% kable() %>%
+ kable_styling(bootstrap_options = "striped", full_width = F)Relationships can also be undirected and weighted
Relationships can also be directed and unweighted
Relationships can also be undirected and
+weighted
Relationships can also be directed and
+unweighted
Chord diagram is a good way to represent the migration flows. It works well if your data are directed and weighted like for migration flows between country.
-Disclaimer: this plot is made using the circlize library, and very strongly inspired from the Migest package from Gui J. Abel, who is also the author of the migration dataset used here.
-Since this kind of graphic is used to display flows, it can be applied only on network where connection are weighted. It does not work for the other example on authors connections.
# short names
-colnames(data) <- c("Africa", "East Asia", "Europe", "Latin Ame.", "North Ame.", "Oceania", "South Asia", "South East Asia", "Soviet Union", "West.Asia")
-rownames(data) <- colnames(data)
-
-# I need a long format
-data_long <- data %>%
- rownames_to_column %>%
- gather(key = 'key', value = 'value', -rowname)
-
-# parameters
-circos.clear()
-circos.par(start.degree = 90, gap.degree = 4, track.margin = c(-0.1, 0.1), points.overflow.warning = FALSE)
-par(mar = rep(0, 4))
-
-# color palette
-mycolor <- viridis(10, alpha = 1, begin = 0, end = 1, option = "D")
-mycolor <- mycolor[sample(1:10)]
-
-# Base plot
-chordDiagram(
- x = data_long,
- grid.col = mycolor,
- transparency = 0.25,
- directional = 1,
- direction.type = c("arrows", "diffHeight"),
- diffHeight = -0.04,
- annotationTrack = "grid",
- annotationTrackHeight = c(0.05, 0.1),
- link.arr.type = "big.arrow",
- link.sort = TRUE,
- link.largest.ontop = TRUE)
-
-# Add text and axis
-circos.trackPlotRegion(
- track.index = 1,
- bg.border = NA,
- panel.fun = function(x, y) {
-
- xlim = get.cell.meta.data("xlim")
- sector.index = get.cell.meta.data("sector.index")
-
- # Add names to the sector.
- circos.text(
- x = mean(xlim),
- y = 3.2,
- labels = sector.index,
- facing = "bending",
- cex = 0.8
- )
-
- # Add graduation on axis
- circos.axis(
- h = "top",
- major.at = seq(from = 0, to = xlim[2], by = ifelse(test = xlim[2]>10, yes = 2, no = 1)),
- minor.ticks = 1,
- major.tick.percentage = 0.5,
- labels.niceFacing = FALSE)
- }
-)
In my opinion this is a powerful way to display information. Major flows are easy to detect, like the migration from South Asia towars Westa Asia, or Africa to Europe. Moreover, for each continent it is quite easy to quantify the proportion of people leaving and arriving.
However chord diagram is not an usual way of displaying information. Thus, it is advised to give a good amount of explanation to educate your audience. A good way to do so is to draw just a few connections in a first step, before displaying the whole graphic. See this blog post by Nadieh Bremer for more ideas on this topic.
-A chord diagram is a good way to represent migration flows. It works +well if your data are directed and weighted like for migration flows +between country.
+Disclaimer: this plot is made using the circlize library, and very +strongly inspired from the Migest package from Gui J. Abel, who is also the author of the +migration dataset +used here.
+Since this kind of graphic is used to display flows, it can be
+applied only on networks in which connections are weighted.
+It does not work for the other example on authors connections.
# short names
+colnames(data) <- c("Africa", "East Asia", "Europe", "Latin Ame.", "North Ame.", "Oceania", "South Asia", "South East Asia", "Soviet Union", "West.Asia")
+rownames(data) <- colnames(data)
+
+# I need a long format
+data_long <- data %>%
+ rownames_to_column %>%
+ gather(key = 'key', value = 'value', -rowname)
+
+# parameters
+circos.clear()
+circos.par(start.degree = 90, gap.degree = 4, track.margin = c(-0.1, 0.1), points.overflow.warning = FALSE)
+par(mar = rep(0, 4))
+
+# color palette
+mycolor <- viridis(10, alpha = 1, begin = 0, end = 1, option = "D")
+mycolor <- mycolor[sample(1:10)]
+
+# Base plot
+chordDiagram(
+ x = data_long,
+ grid.col = mycolor,
+ transparency = 0.25,
+ directional = 1,
+ direction.type = c("arrows", "diffHeight"),
+ diffHeight = -0.04,
+ annotationTrack = "grid",
+ annotationTrackHeight = c(0.05, 0.1),
+ link.arr.type = "big.arrow",
+ link.sort = TRUE,
+ link.largest.ontop = TRUE)
+
+# Add text and axis
+circos.trackPlotRegion(
+ track.index = 1,
+ bg.border = NA,
+ panel.fun = function(x, y) {
+
+ xlim = get.cell.meta.data("xlim")
+ sector.index = get.cell.meta.data("sector.index")
+
+ # Add names to the sector.
+ circos.text(
+ x = mean(xlim),
+ y = 3.2,
+ labels = sector.index,
+ facing = "bending",
+ cex = 0.8
+ )
+
+ # Add graduation on axis
+ circos.axis(
+ h = "top",
+ major.at = seq(from = 0, to = xlim[2], by = ifelse(test = xlim[2]>10, yes = 2, no = 1)),
+ minor.ticks = 1,
+ major.tick.percentage = 0.5,
+ labels.niceFacing = FALSE)
+ }
+)## `major.tick.percentage` is not used any more, please directly use argument `major.tick.length`.
+## `major.tick.percentage` is not used any more, please directly use argument `major.tick.length`.
+## `major.tick.percentage` is not used any more, please directly use argument `major.tick.length`.
+## `major.tick.percentage` is not used any more, please directly use argument `major.tick.length`.
+## `major.tick.percentage` is not used any more, please directly use argument `major.tick.length`.
+## `major.tick.percentage` is not used any more, please directly use argument `major.tick.length`.
+## `major.tick.percentage` is not used any more, please directly use argument `major.tick.length`.
+## `major.tick.percentage` is not used any more, please directly use argument `major.tick.length`.
+## `major.tick.percentage` is not used any more, please directly use argument `major.tick.length`.
+## `major.tick.percentage` is not used any more, please directly use argument `major.tick.length`.
+In my opinion this is a powerful way to display information. Major flows
+are easy to detect, like the migration from South Asia towards Westa
+Asia, or Africa to Europe. Moreover, for each continent it is quite easy
+to quantify the proportion of people leaving and arriving.
However chord diagram is not an usual way of displaying information. +Thus, it is advised to give a good amount of explanation to educate your +audience. A good way to do so is to draw just a few connections in a +first step, before displaying the whole graphic. See this blog post by +Nadieh +Bremer for more ideas on this topic.
+Sankey diagram is another option to display weighted connection. Instead of displaying regions on a circle, they are duplicated and represented on both side of the graphic. Origin is usually on the left, destination on the right.
-# Package
-library(networkD3)
-
-# I need a long format
-data_long <- data %>%
- rownames_to_column %>%
- gather(key = 'key', value = 'value', -rowname) %>%
- filter(value > 0)
-colnames(data_long) <- c("source", "target", "value")
-data_long$target <- paste(data_long$target, " ", sep="")
-
-# From these flows we need to create a node data frame: it lists every entities involved in the flow
-nodes <- data.frame(name=c(as.character(data_long$source), as.character(data_long$target)) %>% unique())
-
-# With networkD3, connection must be provided using id, not using real name like in the links dataframe.. So we need to reformat it.
-data_long$IDsource=match(data_long$source, nodes$name)-1
-data_long$IDtarget=match(data_long$target, nodes$name)-1
-
-# prepare colour scale
-ColourScal ='d3.scaleOrdinal() .range(["#FDE725FF","#B4DE2CFF","#6DCD59FF","#35B779FF","#1F9E89FF","#26828EFF","#31688EFF","#3E4A89FF","#482878FF","#440154FF"])'
-
-# Make the Network
-sankeyNetwork(Links = data_long, Nodes = nodes,
- Source = "IDsource", Target = "IDtarget",
- Value = "value", NodeID = "name",
- sinksRight=FALSE, colourScale=ColourScal, nodeWidth=40, fontSize=13, nodePadding=20)Sankey diagram is another option +to display weighted connection. Intead of displaying regions on a +circle, they are duplicated and represented on both sides of the +graphic. Origin is usually on the left, destination on the right.
+# Package
+library(networkD3)
+
+# I need a long format
+data_long <- data %>%
+ rownames_to_column %>%
+ gather(key = 'key', value = 'value', -rowname) %>%
+ filter(value > 0)
+colnames(data_long) <- c("source", "target", "value")
+data_long$target <- paste(data_long$target, " ", sep="")
+
+# From these flows we need to create a node data frame: it lists every entities involved in the flow
+nodes <- data.frame(name=c(as.character(data_long$source), as.character(data_long$target)) %>% unique())
+
+# With networkD3, connection must be provided using id, not using real name like in the links dataframe.. So we need to reformat it.
+data_long$IDsource=match(data_long$source, nodes$name)-1
+data_long$IDtarget=match(data_long$target, nodes$name)-1
+
+# prepare colour scale
+ColourScal ='d3.scaleOrdinal() .range(["#FDE725FF","#B4DE2CFF","#6DCD59FF","#35B779FF","#1F9E89FF","#26828EFF","#31688EFF","#3E4A89FF","#482878FF","#440154FF"])'
+
+# Make the Network
+sankeyNetwork(Links = data_long, Nodes = nodes,
+ Source = "IDsource", Target = "IDtarget",
+ Value = "value", NodeID = "name",
+ sinksRight=FALSE, colourScale=ColourScal, nodeWidth=40, fontSize=13, nodePadding=20)The heatmap is another great alternative to represent an adjacency matrix. Here, all the origin countries are represented as row, and all the destination as columns. The diagonal pops out with a lot of yellow squares, what means that most of the migrations are intra continental.
-library(heatmaply)
-p <- heatmaply(data,
- dendrogram = "none",
- xlab = "", ylab = "",
- main = "",
- scale = "column",
- margins = c(60,100,40,20),
- grid_color = "white",
- grid_width = 0.00001,
- titleX = FALSE,
- hide_colorbar = TRUE,
- branches_lwd = 0.1,
- label_names = c("From", "To:", "Value"),
- fontsize_row = 7, fontsize_col = 7,
- labCol = colnames(data),
- labRow = rownames(data),
- heatmap_layers = theme(axis.line=element_blank())
- )The heatmap is +another great alternative to represent an adjacency matrix. Here, all +the origin countries are represented as row, and all the destination as +columns. The diagonal pops out with a lot of yellow squares, which means +that most of the migrations are intra continental.
+library(heatmaply)
+p <- heatmaply(data,
+ dendrogram = "none",
+ xlab = "", ylab = "",
+ main = "",
+ scale = "column",
+ margins = c(60,100,40,20),
+ grid_color = "white",
+ grid_width = 0.00001,
+ titleX = FALSE,
+ hide_colorbar = TRUE,
+ branches_lwd = 0.1,
+ label_names = c("From", "To:", "Value"),
+ fontsize_row = 7, fontsize_col = 7,
+ labCol = colnames(data),
+ labRow = rownames(data),
+ heatmap_layers = theme(axis.line=element_blank())
+ )Note that if the matrix is unweighted, each connection can have only 2 values: 1 if there is a connection, 0 otherwise. It is the case for the co-authorship network example, where researchers are connected if they have already published a paper together. The heatmap below shows these connection and also apply a clustering algorithm to the data: researchers that tend to be involved in the same papers are grouped together.
# Format data
-tmp <- dataUU
-rownames(tmp) <- tmp$from
-tmp <- tmp %>% select(-from)
-tmp[is.na(tmp)] <- 0
-
-# Keep people with more than 1 connections
-tmp <- tmp[which(rowSums(tmp)>3), which(colSums(tmp)>3)]
-
-# Heatmap
-p <- heatmaply(tmp,
- dendrogram = "both",
- xlab = "", ylab = "",
- main = "",
- scale = "none",
- margins = c(60,100,40,20),
- grid_color = "white",
- grid_width = 0.0000000001,
- titleX = FALSE,
- hide_colorbar = TRUE,
- branches_lwd = 0.1,
- label_names = c("Name", "With:", "Value"),
- fontsize_row = 7, fontsize_col = 7,
- labCol = colnames(tmp),
- labRow = rownames(tmp),
- heatmap_layers = theme(axis.line=element_blank())
- )Note that if the matrix is unweighted, each connection
+can have only 2 values: 1 if there is a connection, 0 otherwise. It is
+the case for the co-authorship network example, where researchers are
+connected if they have already published a paper together. The heatmap
+below shows these connections and also applies a clustering algorithm to
+the data: researchers that tend to be involved in the same papers are
+grouped together.
# Format data
+tmp <- dataUU
+rownames(tmp) <- tmp$from
+tmp <- tmp %>% select(-from)
+tmp[is.na(tmp)] <- 0
+
+# Keep people with more than 1 connections
+tmp <- tmp[which(rowSums(tmp)>3), which(colSums(tmp)>3)]
+
+# Heatmap
+p <- heatmaply(tmp,
+ dendrogram = "both",
+ xlab = "", ylab = "",
+ main = "",
+ scale = "none",
+ margins = c(60,100,40,20),
+ grid_color = "white",
+ grid_width = 0.0000000001,
+ titleX = FALSE,
+ hide_colorbar = TRUE,
+ branches_lwd = 0.1,
+ label_names = c("Name", "With:", "Value"),
+ fontsize_row = 7, fontsize_col = 7,
+ labCol = colnames(tmp),
+ labRow = rownames(tmp),
+ heatmap_layers = theme(axis.line=element_blank())
+ )Since an adjacency matrix is a network structure, it is possible to build a network graph. In a network graph, each entity is represented as a node, and each connection as an edge.
In my opinion, this type of representation makes more sense when the connection are unweighted, since drawing edges with different sizes tends to clutter the figure and make it unreadable.
Thus, here is an application of this chart type to the coauthor network. Researchers are the nodes, represented as dots. If 2 researchers have published at least one scientific paper together, they are connected. The node size is proportionnal to the number of coauthors.
+Since an adjacency matrix is a network structure, it is
+possible to build a network graph.
+In a network graph, each entity is represented as a node,
+and each connection as an edge.
In my opinion, this type of representation makes more sense when the
+connections are unweighted, since drawing edges with
+different sizes tends to clutter the figure and make it unreadable.
Thus, here is an application of this chart type to the coauthor +network. Researchers are the nodes, represented as dots. If 2 +researchers have published at least one scientific paper together, they +are connected. The node size is proportionnal to the number of +coauthors.
+# Transform the adjacency matrix in a long format
-connect <- dataUU %>%
- gather(key="to", value="value", -1) %>%
- mutate(to = gsub("\\.", " ",to)) %>%
- na.omit()
-
-# Number of connection per person
-c( as.character(connect$from), as.character(connect$to)) %>%
- as.tibble() %>%
- group_by(value) %>%
- summarize(n=n()) -> coauth
-colnames(coauth) <- c("name", "n")
-
-# Create a graph object with igraph
-mygraph <- graph_from_data_frame( connect, vertices = coauth, directed = FALSE )
-
-# Find community
-com <- walktrap.community(mygraph)
-
-#Reorder dataset and make the graph
-coauth <- coauth %>%
- mutate( grp = com$membership) %>%
- arrange(grp) %>%
- mutate(name=factor(name, name))
-
-# keep only 10 first communities
-coauth <- coauth %>%
- filter(grp<16)
-
-# keep only this people in edges
-connect <- connect %>%
- filter(from %in% coauth$name) %>%
- filter(to %in% coauth$name)
-
-# Create a graph object with igraph
-mygraph <- graph_from_data_frame( connect, vertices = coauth, directed = FALSE )
-
-# prepare a vector of n color in the viridis scale
-mycolor <- colormap(colormap=colormaps$viridis, nshades=max(coauth$grp))
-mycolor <- sample(mycolor, length(mycolor))
-
-# Make the graph
-ggraph(mygraph) +
- geom_edge_link(edge_colour="black", edge_alpha=0.2, edge_width=0.3, fold=TRUE) +
- geom_node_point(aes(size=n, color=as.factor(grp), fill=grp), alpha=0.9) +
- scale_size_continuous(range=c(0.5,8)) +
- scale_color_manual(values=mycolor) +
- geom_node_text(aes(label=ifelse(n>6, as.character(name), "")), angle=65, hjust=rep(c(0,1),58), nudge_y = rep(c(0.5,-0.5),58), size=2.3, color="grey") +
- theme_void() +
- theme(
- legend.position="none",
- plot.margin=unit(c(0,0,0,0), "null"),
- panel.spacing=unit(c(0,0,0,0), "null")
- ) +
- expand_limits(x = c(-1.2, 1.2), y = c(-1.2, 1.2)) # Transform the adjacency matrix in a long format
+connect <- dataUU %>%
+ gather(key="to", value="value", -1) %>%
+ mutate(to = gsub("\\.", " ",to)) %>%
+ na.omit()
+
+# Number of connection per person
+c( as.character(connect$from), as.character(connect$to)) %>%
+ as.tibble() %>%
+ group_by(value) %>%
+ summarize(n=n()) -> coauth
+colnames(coauth) <- c("name", "n")
+
+# Create a graph object with igraph
+mygraph <- graph_from_data_frame( connect, vertices = coauth, directed = FALSE )
+
+# Find community
+com <- walktrap.community(mygraph)
+
+#Reorder dataset and make the graph
+coauth <- coauth %>%
+ mutate( grp = com$membership) %>%
+ arrange(grp) %>%
+ mutate(name=factor(name, name))
+
+# keep only 10 first communities
+coauth <- coauth %>%
+ filter(grp<16)
+
+# keep only this people in edges
+connect <- connect %>%
+ filter(from %in% coauth$name) %>%
+ filter(to %in% coauth$name)
+
+# Create a graph object with igraph
+mygraph <- graph_from_data_frame( connect, vertices = coauth, directed = FALSE )
+
+# prepare a vector of n color in the viridis scale
+mycolor <- colormap(colormap=colormaps$viridis, nshades=max(coauth$grp))
+mycolor <- sample(mycolor, length(mycolor))
+
+# Make the graph
+ggraph(mygraph) +
+ geom_edge_link(edge_colour="black", edge_alpha=0.2, edge_width=0.3, fold=TRUE) +
+ geom_node_point(aes(size=n, color=as.factor(grp), fill=grp), alpha=0.9) +
+ scale_size_continuous(range=c(0.5,8)) +
+ scale_color_manual(values=mycolor) +
+ geom_node_text(aes(label=ifelse(n>6, as.character(name), "")), angle=65, hjust=rep(c(0,1),58), nudge_y = rep(c(0.5,-0.5),58), size=2.3, color="grey") +
+ theme_void() +
+ theme(
+ legend.position="none",
+ plot.margin=unit(c(0,0,0,0), "null"),
+ panel.spacing=unit(c(0,0,0,0), "null")
+ ) +
+ expand_limits(x = c(-1.2, 1.2), y = c(-1.2, 1.2))
Network graphs are very powerful to study the global structure of the network. Here, a few groups of researchers are isolated. Each actually represents one single paper where Vincent Ranwez was involved. In the middle a massive network of researchers appear: these are the people with Vincent published more often that are thus all linked together.
+Network +graphs are very powerful to study the global structure of the +network. Here, a few groups of researchers are isolated. Each actually +represents one single paper where Vincent Ranwez was involved. In the +middle a massive network of researchers appear: these are the people who +Vincent published with most often, and are therefore all linked +together.
However, network charts are very bad a annotating every single points: names tend to overlap edges making the figure unreadable. The arc diagram described below is a good alternative if you want to show labels.
-However, network charts are very bad at annotating every single +points: names tend to overlap edges making the figure unreadable. The +arc diagram described below is a good alternative if you want to show +labels.
+Instead of using a custom algorithm to position each nodes, it is possible to place them all around a circule, making a chord diagram. But this kind of chart makes sense only if the order of nodes around the circule is carefully chosen, to avoid having a cluttered and unreadable figure.
-# Transform the adjacency matrix in a long format
-connect <- dataUU %>%
- gather(key="to", value="value", -1) %>%
- mutate(to = gsub("\\.", " ",to)) %>%
- na.omit()
-
-# Number of connection per person
-c( as.character(connect$from), as.character(connect$to)) %>%
- as.tibble() %>%
- group_by(value) %>%
- summarize(n=n()) -> coauth
-colnames(coauth) <- c("name", "n")
-
-# Create a graph object with igraph
-mygraph <- graph_from_data_frame( connect, vertices = coauth, directed = FALSE )
-
-# Find community
-com <- walktrap.community(mygraph)
-
-#Reorder dataset and make the graph
-coauth <- coauth %>%
- mutate( grp = com$membership) %>%
- arrange(grp) %>%
- mutate(name=factor(name, name))
-
-# keep only 10 first communities
-coauth <- coauth %>%
- filter(grp<16)
-
-# keep only this people in edges
-connect <- connect %>%
- filter(from %in% coauth$name) %>%
- filter(to %in% coauth$name)
-
-# Add label angle
-number_of_bar=nrow(coauth)
-coauth$id = seq(1, nrow(coauth))
-angle= 360 * (coauth$id-0.5) /number_of_bar # I substract 0.5 because the letter must have the angle of the center of the bars. Not extreme right(1) or extreme left (0)
-coauth$hjust <- ifelse(angle > 90 & angle<270, 1, 0)
-coauth$angle <- ifelse(angle > 90 & angle<270, angle+180, angle)
-
-# Create a graph object with igraph
-mygraph <- graph_from_data_frame( connect, vertices = coauth, directed = FALSE )
-
-# prepare a vector of n color in the viridis scale
-mycolor <- colormap(colormap=colormaps$viridis, nshades=max(coauth$grp))
-mycolor <- sample(mycolor, length(mycolor))
-
-# Make the graph
-ggraph(mygraph, layout="circle") +
- geom_edge_link(edge_colour="black", edge_alpha=0.2, edge_width=0.3, fold=FALSE) +
- geom_node_point(aes(size=n, color=as.factor(grp), fill=grp), alpha=0.9) +
- scale_size_continuous(range=c(0.5,8)) +
- scale_color_manual(values=mycolor) +
- geom_node_text(aes(label=paste(" ",name," "), angle=angle, hjust=hjust), size=2.3, color="black") +
- theme_void() +
- theme(
- legend.position="none",
- plot.margin=unit(c(0,0,0,0), "null"),
- panel.spacing=unit(c(0,0,0,0), "null")
- ) +
- expand_limits(x = c(-1.2, 1.2), y = c(-1.2, 1.2)) Instead of using a custom algorithm to position each nodes, it is +possible to place them all around a circle, making a chord diagram. But +this kind of chart makes sense only if the order of nodes around the +circle is carefully chosen, to avoid having a cluttered and unreadable +figure.
+# Transform the adjacency matrix in a long format
+connect <- dataUU %>%
+ gather(key="to", value="value", -1) %>%
+ mutate(to = gsub("\\.", " ",to)) %>%
+ na.omit()
+
+# Number of connection per person
+c( as.character(connect$from), as.character(connect$to)) %>%
+ as.tibble() %>%
+ group_by(value) %>%
+ summarize(n=n()) -> coauth
+colnames(coauth) <- c("name", "n")
+
+# Create a graph object with igraph
+mygraph <- graph_from_data_frame( connect, vertices = coauth, directed = FALSE )
+
+# Find community
+com <- walktrap.community(mygraph)
+
+#Reorder dataset and make the graph
+coauth <- coauth %>%
+ mutate( grp = com$membership) %>%
+ arrange(grp) %>%
+ mutate(name=factor(name, name))
+
+# keep only 10 first communities
+coauth <- coauth %>%
+ filter(grp<16)
+
+# keep only this people in edges
+connect <- connect %>%
+ filter(from %in% coauth$name) %>%
+ filter(to %in% coauth$name)
+
+# Add label angle
+number_of_bar=nrow(coauth)
+coauth$id = seq(1, nrow(coauth))
+angle= 360 * (coauth$id-0.5) /number_of_bar # I substract 0.5 because the letter must have the angle of the center of the bars. Not extreme right(1) or extreme left (0)
+coauth$hjust <- ifelse(angle > 90 & angle<270, 1, 0)
+coauth$angle <- ifelse(angle > 90 & angle<270, angle+180, angle)
+
+# Create a graph object with igraph
+mygraph <- graph_from_data_frame( connect, vertices = coauth, directed = FALSE )
+
+# prepare a vector of n color in the viridis scale
+mycolor <- colormap(colormap=colormaps$viridis, nshades=max(coauth$grp))
+mycolor <- sample(mycolor, length(mycolor))
+
+# Make the graph
+ggraph(mygraph, layout="circle") +
+ geom_edge_link(edge_colour="black", edge_alpha=0.2, edge_width=0.3, fold=FALSE) +
+ geom_node_point(aes(size=n, color=as.factor(grp), fill=grp), alpha=0.9) +
+ scale_size_continuous(range=c(0.5,8)) +
+ scale_color_manual(values=mycolor) +
+ geom_node_text(aes(label=paste(" ",name," "), angle=angle, hjust=hjust), size=2.3, color="black") +
+ theme_void() +
+ theme(
+ legend.position="none",
+ plot.margin=unit(c(0,0,0,0), "null"),
+ panel.spacing=unit(c(0,0,0,0), "null")
+ ) +
+ expand_limits(x = c(-1.2, 1.2), y = c(-1.2, 1.2))
An arc diagram follows the same concept, but displays nodes along a single axis and links with arcs. The main advantage is that it allows to make the labels easy to read.
+An arc diagram follows the same concept, but displays nodes along a +single axis and links with arcs. The main advantage is that it allows to +make the labels easy to read.
+# Make the graph
-ggraph(mygraph, layout="linear") +
- geom_edge_arc(edge_colour="black", edge_alpha=0.2, edge_width=0.3, fold=TRUE) +
- geom_node_point(aes(size=n, color=as.factor(grp), fill=grp), alpha=0.5) +
- scale_size_continuous(range=c(0.5,8)) +
- scale_color_manual(values=mycolor) +
- geom_node_text(aes(label=name), angle=65, hjust=1, nudge_y = -1.1, size=2.3) +
- theme_void() +
- theme(
- legend.position="none",
- plot.margin=unit(c(0,0,0.4,0), "null"),
- panel.spacing=unit(c(0,0,3.4,0), "null")
- ) +
- expand_limits(x = c(-1.2, 1.2), y = c(-5.6, 1.2)) # Make the graph
+ggraph(mygraph, layout="linear") +
+ geom_edge_arc(edge_colour="black", edge_alpha=0.2, edge_width=0.3, fold=TRUE) +
+ geom_node_point(aes(size=n, color=as.factor(grp), fill=grp), alpha=0.5) +
+ scale_size_continuous(range=c(0.5,8)) +
+ scale_color_manual(values=mycolor) +
+ geom_node_text(aes(label=name), angle=65, hjust=1, nudge_y = -1.1, size=2.3) +
+ theme_void() +
+ theme(
+ legend.position="none",
+ plot.margin=unit(c(0,0,0.4,0), "null"),
+ panel.spacing=unit(c(0,0,3.4,0), "null")
+ ) +
+ expand_limits(x = c(-1.2, 1.2), y = c(-5.6, 1.2))
You can learn more about each type of graphic presented in this story in the dedicated sections. Click the icon below:
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
- Any thoughts on this? Found any mistake? Have another way to show the data? Please drop me a word on Twitter or in the comment section below:
-+ You can learn more about each type of graphic presented in this + story in the dedicated sections. Click the icon below: +
+ +
+
+
+
+
+
+
+
+
+ + Data To Viz is a + comprehensive classification of chart types organized by data + input format. Get a high-resolution version of our decision tree + delivered to your inbox now! +
+
+ A work by Yan Holtz for data-to-viz.com
@@ -789,28 +984,28 @@
- The biggest UK cities
-
-
-
-
-
-
-
-
-
-
-
-
+ The biggest UK cities
+
+
This document gives a few suggestions to analyse a dataset composed by a set of geographic coordinates that have an associated numeric value.
It considers the population of 925 cities in the UK. This example dataset is provided in the R maps library and is available on this Github repository. Basically it looks like the table beside.
This document gives a few suggestions to analyse a dataset composed
+by a set of geographic coordinates that have an associated numeric
+value.
It considers the population of 925 cities in the UK. This
+example dataset is provided in the R maps library and is
+available on this Github
+repository. Basically it looks like the table beside.
# Libraries
-library(tidyverse)
-library(hrbrthemes)
-library(kableExtra)
-options(knitr.table.format = "html")
-library(viridis)
-
-# Load dataset from github
-data <- read.table("https://raw.githubusercontent.com/holtzy/data_to_viz/master/Example_dataset/18_ListGPSCoordinatesWithValue.csv", header=T)
-
-# show data
-data %>% head(5) %>% kable() %>%
- kable_styling(bootstrap_options = "striped", full_width = F)# Libraries
+library(tidyverse)
+library(hrbrthemes)
+library(kableExtra)
+options(knitr.table.format = "html")
+library(viridis)
+
+# Load dataset from github
+data <- read.table("https://raw.githubusercontent.com/holtzy/data_to_viz/master/Example_dataset/18_ListGPSCoordinatesWithValue.csv", header=T)
+
+# show data
+data %>% head(5) %>% kable() %>%
+ kable_styling(bootstrap_options = "striped", full_width = F)If your dataset is composed by a few data point only, you can just display them on a map. If you have specific information to display concerning these positions, use an interactive: more information are available when you click data points.
-If your dataset is composed by a few data point only, you can just +display them on a map. If you have specific information to display +concerning these positions, use an interactive: more information are +available when you click data points.
+A dot density map is used when the sample size of your dataset is high.
-# Get the world polygon
-world <- map_data("world")
-
-p <- data %>%
- #head(1000) %>%
- ggplot( aes(x=homelon, y=homelat)) +
- geom_polygon(data = world, aes(x=long, y = lat, group = group), fill="grey", alpha=0.1) +
- geom_point(size=0.8, color="#69b3a2", alpha=0.5) +
- coord_equal() +
- theme_void() +
- theme(
- panel.spacing=unit(c(0,0,0,0), "null"),
- plot.margin=grid::unit(c(0,0,0,0), "cm"),
- ) +
- ggplot2::annotate("text", x = -150, y = -45, hjust = 0, size = 11, label = paste("Where surfers live."), color = "Black") +
- ggplot2::annotate("text", x = -150, y = -51, hjust = 0, size = 8, label = paste("data-to-viz.com | 200,000 #surf tweets recovered"), color = "black", alpha = 0.5) +
- xlim(-180,180) +
- ylim(-60,80) +
- scale_x_continuous(expand = c(0.006, 0.006)) +
- coord_equal()
-
-ggsave(p, file="IMG/Surfer_position.png", width = 36, height = 15.22, units = "in", dpi = 90)A dot density map is used when the sample size of your dataset is +high.
+# Get the world polygon
+world <- map_data("world")
+
+p <- data %>%
+ #head(1000) %>%
+ ggplot( aes(x=homelon, y=homelat)) +
+ geom_polygon(data = world, aes(x=long, y = lat, group = group), fill="grey", alpha=0.1) +
+ geom_point(size=0.8, color="#69b3a2", alpha=0.5) +
+ coord_equal() +
+ theme_void() +
+ theme(
+ panel.spacing=unit(c(0,0,0,0), "null"),
+ plot.margin=grid::unit(c(0,0,0,0), "cm"),
+ ) +
+ ggplot2::annotate("text", x = -150, y = -45, hjust = 0, size = 11, label = paste("Where surfers live."), color = "Black") +
+ ggplot2::annotate("text", x = -150, y = -51, hjust = 0, size = 8, label = paste("data-to-viz.com | 200,000 #surf tweets recovered"), color = "black", alpha = 0.5) +
+ xlim(-180,180) +
+ ylim(-60,80) +
+ scale_x_continuous(expand = c(0.006, 0.006)) +
+ coord_equal()
+
+ggsave(p, file="IMG/Surfer_position.png", width = 36, height = 15.22, units = "in", dpi = 90)
-
Dotmaps give a good idea about where samples are distributed. However, once dots start to overlap, it gets impossible to distinguish how many of them are displayed on a certain zone. That’s where binning becomes an interesting option.
-Dotmaps give a good idea about where samples are distributed. +However, once dots start to overlap, it gets impossible to distinguish +how many of them are displayed on a certain zone. That’s where binning +becomes an interesting option.
+To create a hexbin map, the territory is divided in many hexagons and the number of sample per hexagon is counted and represented by a color.
To create a hexbin map,
+the territory is divided in many hexagons and the number of
+sample per hexagon is counted and represented by a color.
data %>%
- filter(homecontinent=='Europe') %>%
- ggplot( aes(x=homelon, y=homelat)) +
- geom_hex(bins=59) +
- ggplot2::annotate("text", x = -27, y = 72, label="Where people tweet about #Surf", colour = "black", size=5, alpha=1, hjust=0) +
- ggplot2::annotate("segment", x = -27, xend = 10, y = 70, yend = 70, colour = "black", size=0.2, alpha=1) +
- theme_void() +
- xlim(-30, 70) +
- ylim(24, 72) +
- scale_fill_viridis(
- trans = "log",
- breaks = c(1,7,54,403,3000),
- name="Tweet # recorded in 8 months",
- guide = guide_legend( keyheight = unit(2.5, units = "mm"), keywidth=unit(10, units = "mm"), label.position = "bottom", title.position = 'top', nrow=1)
- ) +
- ggtitle( "" ) +
- theme(
- legend.position = c(0.8, 0.09),
- legend.title=element_text(color="black", size=8),
- text = element_text(color = "#22211d"),
- plot.background = element_rect(fill = "#f5f5f2", color = NA),
- panel.background = element_rect(fill = "#f5f5f2", color = NA),
- legend.background = element_rect(fill = "#f5f5f2", color = NA),
- plot.title = element_text(size= 13, hjust=0.1, color = "#4e4d47", margin = margin(b = -0.1, t = 0.4, l = 2, unit = "cm")),
- ) data %>%
+ filter(homecontinent=='Europe') %>%
+ ggplot( aes(x=homelon, y=homelat)) +
+ geom_hex(bins=59) +
+ ggplot2::annotate("text", x = -27, y = 72, label="Where people tweet about #Surf", colour = "black", size=5, alpha=1, hjust=0) +
+ ggplot2::annotate("segment", x = -27, xend = 10, y = 70, yend = 70, colour = "black", size=0.2, alpha=1) +
+ theme_void() +
+ xlim(-30, 70) +
+ ylim(24, 72) +
+ scale_fill_viridis(
+ trans = "log",
+ breaks = c(1,7,54,403,3000),
+ name="Tweet # recorded in 8 months",
+ guide = guide_legend( keyheight = unit(2.5, units = "mm"), keywidth=unit(10, units = "mm"), label.position = "bottom", title.position = 'top', nrow=1)
+ ) +
+ ggtitle( "" ) +
+ theme(
+ legend.position = c(0.8, 0.09),
+ legend.title=element_text(color="black", size=8),
+ text = element_text(color = "#22211d"),
+ plot.background = element_rect(fill = "#f5f5f2", color = NA),
+ panel.background = element_rect(fill = "#f5f5f2", color = NA),
+ legend.background = element_rect(fill = "#f5f5f2", color = NA),
+ plot.title = element_text(size= 13, hjust=0.1, color = "#4e4d47", margin = margin(b = -0.1, t = 0.4, l = 2, unit = "cm")),
+ )
Note that this is very close from an 2d histogram map. Basically, it splits the space into a set of squares instead of hexagons, and uses the same process:
+Note that this is very close from an 2d histogram +map. Basically, it splits the space into a set of squares instead of +hexagons, and uses the same process:
# Make the hexbin map with the geom_hex function
-ggplot(data, aes(x=homelon, y=homelat)) +
- geom_polygon(data = world, aes(x=long, y = lat, group = group), fill="grey", alpha=0.3) +
- geom_bin2d(bins=100) +
- ggplot2::annotate("text", x = 175, y = 80, label="Where people tweet about #Surf", colour = "black", size=4, alpha=1, hjust=1) +
- ggplot2::annotate("segment", x = 100, xend = 175, y = 73, yend = 73, colour = "black", size=0.2, alpha=1) +
- theme_void() +
- ylim(-70, 80) +
- scale_fill_viridis(
- trans = "log",
- breaks = c(1,7,54,403,3000),
- name="Tweet # recorded in 8 months",
- guide = guide_legend( keyheight = unit(2.5, units = "mm"), keywidth=unit(10, units = "mm"), label.position = "bottom", title.position = 'top', nrow=1)
- ) +
- ggtitle( "" ) +
- theme(
- legend.position = c(0.8, 0.09),
- legend.title=element_text(color="black", size=8),
- text = element_text(color = "#22211d"),
- plot.background = element_rect(fill = "#f5f5f2", color = NA),
- panel.background = element_rect(fill = "#f5f5f2", color = NA),
- legend.background = element_rect(fill = "#f5f5f2", color = NA),
- plot.title = element_text(size= 13, hjust=0.1, color = "#4e4d47", margin = margin(b = -0.1, t = 0.4, l = 2, unit = "cm")),
- ) # Make the hexbin map with the geom_hex function
+ggplot(data, aes(x=homelon, y=homelat)) +
+ geom_polygon(data = world, aes(x=long, y = lat, group = group), fill="grey", alpha=0.3) +
+ geom_bin2d(bins=100) +
+ ggplot2::annotate("text", x = 175, y = 80, label="Where people tweet about #Surf", colour = "black", size=4, alpha=1, hjust=1) +
+ ggplot2::annotate("segment", x = 100, xend = 175, y = 73, yend = 73, colour = "black", size=0.2, alpha=1) +
+ theme_void() +
+ ylim(-70, 80) +
+ scale_fill_viridis(
+ trans = "log",
+ breaks = c(1,7,54,403,3000),
+ name="Tweet # recorded in 8 months",
+ guide = guide_legend( keyheight = unit(2.5, units = "mm"), keywidth=unit(10, units = "mm"), label.position = "bottom", title.position = 'top', nrow=1)
+ ) +
+ ggtitle( "" ) +
+ theme(
+ legend.position = c(0.8, 0.09),
+ legend.title=element_text(color="black", size=8),
+ text = element_text(color = "#22211d"),
+ plot.background = element_rect(fill = "#f5f5f2", color = NA),
+ panel.background = element_rect(fill = "#f5f5f2", color = NA),
+ legend.background = element_rect(fill = "#f5f5f2", color = NA),
+ plot.title = element_text(size= 13, hjust=0.1, color = "#4e4d47", margin = margin(b = -0.1, t = 0.4, l = 2, unit = "cm")),
+ )
It is also possible to divide your territory (the world here) in regions other than square or hexagones. This will result in a chloropleth map. Of course, you need the information of the exact shape of your regions.
+It is also possible to divide your territory (the world here) in +regions other than square or hexagones. This will result in a +chloropleth map. Of course, you need the information of the exact shape +of your regions.
You can learn more about each type of graphic presented in this story in the dedicated sections. Click the icon below:
-
-
- 
-
- 
-
-
-
-
-
-
-
-
-
-
-
-
- Any thoughts on this? Found any mistake? Have another way to show the data? Please drop me a word on Twitter or in the comment section below:
-+ You can learn more about each type of graphic presented in this + story in the dedicated sections. Click the icon below: +
+ +
+
+
+
+
+ + Data To Viz is a + comprehensive classification of chart types organized by data + input format. Get a high-resolution version of our decision tree + delivered to your inbox now! +
+
+ A work by Yan Holtz for data-to-viz.com
@@ -478,28 +561,28 @@
- Visualizing geographic connections
-
-
-
-
-
-
-
-
-
-
-
-
+ Visualizing geographic connections
+
+ This document provides a few suggestions for the visualization of geographical connections.
@@ -457,7 +447,9 @@#Showing a few connection on a map ***
+Before showing all the relationships provided in this dataset, it is important to understand how to visualize a unique connection on a map. It is a common practice to link 2 points using the shortest route @@ -506,7 +498,10 @@ points(x=don$long, y=don$lat, col="#69b3a2", cex=2, pch=20) text(rownames(don), x=don$long, y=don$lat, col="black", cex=1, pos=4)
#Showing many connections on a map ***
+It is then possible to use the same method for the whole dataset composed of about 10,000 connections. With such a sample size, it makes sense to group the connections that have exactly the same starting and @@ -581,69 +576,88 @@ # Save at PNG ggsave("IMG/Surfer_travel.png", width = 36, height = 15.22, units = "in", dpi = 90)
Please note that this map is available here if needed. Even if a connecting map is probably the best option for plotting this kind of dataset, please note that other representation like chord diagrams or networks could make a good job as well.
You can learn more about each type of graphic presented in this story in the dedicated sections. Click the icon below:
-
-
-
-
-
-
-
-
-
-
-
+Any thoughts on this? Found any mistake? Have another way to show the data? Please drop me a word on Twitter or in the comment section below:
-+ You can learn more about each type of graphic presented in this + story in the dedicated sections. Click the icon below: +
+ +
+ + Data To Viz is a + comprehensive classification of chart types organized by data + input format. Get a high-resolution version of our decision tree + delivered to your inbox now! +
+
+ A work by Yan Holtz for data-to-viz.com
@@ -677,26 +691,17 @@
- Nested Several values
-
-
- 
-
-
-
-
-
-
-
-
-
-NEED DATA
-
This document gives a few suggestions to analyse a dataset composed by a unique numeric variable.
It considers the night price of about 10,000 Airbnb appartements on the French Riviera in France.
This example dataset has been downloaded from the Airbnb website and is available on this Github repository. Basically it looks like the table beside.
# Libraries
-library(tidyverse)
-library(hrbrthemes)
-library(kableExtra)
-options(knitr.table.format = "html")
-
-# Load dataset from github
-data <- read.table("https://raw.githubusercontent.com/holtzy/data_to_viz/master/Example_dataset/1_OneNum.csv", header=TRUE)
-
-# show data
-data %>% head(6) %>% kable() %>%
- kable_styling(bootstrap_options = "striped", full_width = F)| -price - | -
|---|
| -75 - | -
| -104 - | -
| -369 - | -
| -300 - | -
| -92 - | -
| -64 - | -
The mos
-You can learn more about each type of graphic presented in this story in the dedicated sections. Click the icon below:
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
- Any thoughts on this? Found any mistake? Have another way to show the data? Please drop me a word on Twitter or in the comment section below:
-A work by Yan Holtz for data-to-viz.com
- - - - - - - - - - - - - - - - - -
- Visualizing a set of geographic coordinates
-
-
-
-
-
-
-
-
-
-
-
-
-
-
+ Visualizing a set of geographic coordinates
+
+
+
+
+
+
This document gives a few suggestions to analyse a dataset composed by a list of GPS coordinates
It considers the geographic position of about 200k tweets containing the hashtags #surf, #windsurf or #kitesurf.
This dataset has been built harvesting twitter every day during about 300 days. It is fully available in this Github repository. Basically it looks like the table beside.
This document gives a few suggestions to analyse a
+dataset composed by a list of GPS coordinates
It considers the
+geographic position of about 200k tweets containing the hashtags
+#surf, #windsurf or
+#kitesurf.
This dataset has been built harvesting
+twitter every day during about 300 days. It is fully available in this
+Github
+repository. Basically it looks like the table beside.
# Libraries
-library(tidyverse)
-library(viridis)
-library(hrbrthemes)
-library(kableExtra)
-options(knitr.table.format = "html")
-library(mapdata)
-
-# Load dataset from github
-#data <- read.table("https://raw.githubusercontent.com/holtzy/data_to_viz/master/Example_dataset/17_ListGPSCoordinates.csv", sep=",", header=T)
-data <- read.table("../Example_dataset/17_ListGPSCoordinates.csv", sep=",", header=T)
-
-# show data
-data %>% head(3) %>% kable() %>%
- kable_styling(bootstrap_options = "striped", full_width = F)# Libraries
+library(tidyverse)
+library(viridis)
+library(hrbrthemes)
+library(kableExtra)
+options(knitr.table.format = "html")
+library(mapdata)
+
+# Load dataset from github
+#data <- read.table("https://raw.githubusercontent.com/holtzy/data_to_viz/master/Example_dataset/17_ListGPSCoordinates.csv", sep=",", header=T)
+data <- read.table("../Example_dataset/17_ListGPSCoordinates.csv", sep=",", header=T)
+
+# show data
+data %>% head(3) %>% kable() %>%
+ kable_styling(bootstrap_options = "striped", full_width = F)If your dataset is composed by a few data point only, you can just display them on a map. If you have specific information to display concerning these positions, use an interactive: more information are available when you click data points.
-If your dataset is composed by a few data point only, you can just +display them on a map. If you have specific information to display +concerning these positions, use an interactive: more information are +available when you click data points.
+A dot density map is used when the sample size of your dataset is high.
-# Get the world polygon
-world <- map_data("world")
-
-data %>%
- #head(1000) %>%
- ggplot( aes(x=homelon, y=homelat)) +
- geom_polygon(data = world, aes(x=long, y = lat, group = group), fill="grey", alpha=0.1) +
- geom_point(size=0.8, color="#69b3a2", alpha=0.5) +
- coord_equal() +
- theme_void() +
- theme(
- panel.spacing=unit(c(0,0,0,0), "null"),
- plot.margin=grid::unit(c(0,0,0,0), "cm"),
- ) +
- ggplot2::annotate("text", x = -150, y = -45, hjust = 0, size = 11, label = paste("Where surfers live."), color = "Black") +
- ggplot2::annotate("text", x = -150, y = -51, hjust = 0, size = 8, label = paste("data-to-viz.com | 200,000 #surf tweets recovered"), color = "black", alpha = 0.5) +
- xlim(-180,180) +
- ylim(-60,80) +
- scale_x_continuous(expand = c(0.006, 0.006)) +
- coord_equal() A dot density map is used when the sample size of your dataset is +high.
+# Get the world polygon
+world <- map_data("world")
+
+data %>%
+ #head(1000) %>%
+ ggplot( aes(x=homelon, y=homelat)) +
+ geom_polygon(data = world, aes(x=long, y = lat, group = group), fill="grey", alpha=0.1) +
+ geom_point(size=0.8, color="#69b3a2", alpha=0.5) +
+ coord_equal() +
+ theme_void() +
+ theme(
+ panel.spacing=unit(c(0,0,0,0), "null"),
+ plot.margin=grid::unit(c(0,0,0,0), "cm"),
+ ) +
+ ggplot2::annotate("text", x = -150, y = -45, hjust = 0, size = 11, label = paste("Where surfers live."), color = "Black") +
+ ggplot2::annotate("text", x = -150, y = -51, hjust = 0, size = 8, label = paste("data-to-viz.com | 200,000 #surf tweets recovered"), color = "black", alpha = 0.5) +
+ xlim(-180,180) +
+ ylim(-60,80) +
+ scale_x_continuous(expand = c(0.006, 0.006)) +
+ coord_equal() 
ggsave("IMG/Surfer_position.png", width = 36, height = 15.22, units = "in", dpi = 90)
Dotmaps give a good idea about where samples are distributed. However, once dots start to overlap, it gets impossible to distinguish how many of them are displayed on a certain zone. That’s where binning becomes an interesting option.
+Dotmaps give a good idea about where samples are distributed. +However, once dots start to overlap, it gets impossible to distinguish +how many of them are displayed on a certain zone. That’s where binning +becomes an interesting option.
To create a hexbin map, the territory is divided in many hexagones and the number of sample per hexagone is counted and represented by a color.
-data %>%
- filter(homecontinent=='Europe') %>%
- ggplot( aes(x=homelon, y=homelat)) +
- geom_hex(bins=65) +
- theme_void() +
- xlim(-30, 70) +
- ylim(24, 72) +
- scale_fill_viridis(
- trans = "log",
- name="Number of Tweet recorded in 8 months",
- guide = guide_legend( keyheight = unit(3, units = "mm"), keywidth=unit(12, units = "mm"), label.position = "bottom", title.position = 'top', nrow=1)
- ) +
- ggtitle( "Where people tweet about #Surf" ) +
- theme(
- legend.position = c(0.5, 0.09),
- text = element_text(color = "#22211d"),
- #plot.background = element_rect(fill = "#f5f5f2", color = NA),
- #panel.background = element_rect(fill = "#f5f5f2", color = NA),
- #legend.background = element_rect(fill = "#f5f5f2", color = NA),
- plot.title = element_text(size= 22, hjust=0.1, color = "#4e4d47", margin = margin(b = -0.1, t = 0.4, l = 2, unit = "cm")),
- )To create a hexbin map, the territory is divided in many hexagones +and the number of sample per hexagone is counted and represented by a +color.
+data %>%
+ filter(homecontinent=='Europe') %>%
+ ggplot( aes(x=homelon, y=homelat)) +
+ geom_hex(bins=65) +
+ theme_void() +
+ xlim(-30, 70) +
+ ylim(24, 72) +
+ scale_fill_viridis(
+ trans = "log",
+ name="Number of Tweet recorded in 8 months",
+ guide = guide_legend( keyheight = unit(3, units = "mm"), keywidth=unit(12, units = "mm"), label.position = "bottom", title.position = 'top', nrow=1)
+ ) +
+ ggtitle( "Where people tweet about #Surf" ) +
+ theme(
+ legend.position = c(0.5, 0.09),
+ text = element_text(color = "#22211d"),
+ #plot.background = element_rect(fill = "#f5f5f2", color = NA),
+ #panel.background = element_rect(fill = "#f5f5f2", color = NA),
+ #legend.background = element_rect(fill = "#f5f5f2", color = NA),
+ plot.title = element_text(size= 22, hjust=0.1, color = "#4e4d47", margin = margin(b = -0.1, t = 0.4, l = 2, unit = "cm")),
+ )
# Make the hexbin map with the geom_hex function
-ggplot(data, aes(x=homelon, y=homelat)) +
- geom_polygon(data = world, aes(x=long, y = lat, group = group), fill="grey", alpha=0.3) +
- geom_bin2d(bins=120) +
- theme_void() +
- scale_fill_viridis(
- trans = "log",
- name="Number of Tweets",
- guide = guide_legend( keyheight = unit(3, units = "mm"), keywidth=unit(12, units = "mm"), label.position = "bottom", title.position = 'top', nrow=1)
- ) +
- theme(
- legend.position = c(0.5, 0.09),
- ) +
- ylim(-70, 80)# Make the hexbin map with the geom_hex function
+ggplot(data, aes(x=homelon, y=homelat)) +
+ geom_polygon(data = world, aes(x=long, y = lat, group = group), fill="grey", alpha=0.3) +
+ geom_bin2d(bins=120) +
+ theme_void() +
+ scale_fill_viridis(
+ trans = "log",
+ name="Number of Tweets",
+ guide = guide_legend( keyheight = unit(3, units = "mm"), keywidth=unit(12, units = "mm"), label.position = "bottom", title.position = 'top', nrow=1)
+ ) +
+ theme(
+ legend.position = c(0.5, 0.09),
+ ) +
+ ylim(-70, 80)
It is also possible to divide your territory (the world here) in regions other than square or hexagones. This will result in a chloropleth map. Of course, you need the information of the exact shape of your regions.
+It is also possible to divide your territory (the world here) in +regions other than square or hexagones. This will result in a +chloropleth map. Of course, you need the information of the exact shape +of your regions.
You can learn more about each type of graphic presented in this story in the dedicated sections. Click the icon below:
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
- Any thoughts on this? Found any mistake? Have another way to show the data? Please drop me a word on Twitter or in the comment section below:
-+ You can learn more about each type of graphic presented in this + story in the dedicated sections. Click the icon below: +
+ +
+
+
+
+
+ + Data To Viz is a + comprehensive classification of chart types organized by data + input format. Get a high-resolution version of our decision tree + delivered to your inbox now! +
+
+ A work by Yan Holtz for data-to-viz.com
@@ -464,28 +539,28 @@
- Visualizing a set of geographic coordinates
-
-
-
-
-
-
-
-
-
-
-
-
-
-
+ Visualizing a set of geographic coordinates
+
+
+
+
+
+
This document gives a few suggestions to analyse a dataset composed by a list of GPS coordinates
It considers the geographic position of about 200k tweets containing the hashtags #surf, #windsurf or #kitesurf.
This dataset has been built harvesting twitter every day during about 300 days. It is fully available in this Github repository. Basically it looks like the table beside.
This document gives a few suggestions to analyse a
+dataset composed by a list of GPS coordinates
It considers the
+geographic position of about 200k tweets containing the hashtags
+#surf, #windsurf or
+#kitesurf.
This dataset has been built harvesting
+twitter every day during about 300 days. It is fully available in this
+Github
+repository. Basically it looks like the table beside.
# Libraries
-library(tidyverse)
-library(viridis)
-library(hrbrthemes)
-library(kableExtra)
-options(knitr.table.format = "html")
-library(mapdata)
-
-# Load dataset from github
-#data <- read.table("https://raw.githubusercontent.com/holtzy/data_to_viz/master/Example_dataset/17_ListGPSCoordinates.csv", sep=",", header=T)
-data <- read.table("../Example_dataset/17_ListGPSCoordinates.csv", sep=",", header=T)
-
-# show data
-data %>% head(3) %>% kable() %>%
- kable_styling(bootstrap_options = "striped", full_width = F)# Libraries
+library(tidyverse)
+library(viridis)
+library(hrbrthemes)
+library(kableExtra)
+options(knitr.table.format = "html")
+library(mapdata)
+
+# Load dataset from github
+#data <- read.table("https://raw.githubusercontent.com/holtzy/data_to_viz/master/Example_dataset/17_ListGPSCoordinates.csv", sep=",", header=T)
+data <- read.table("../Example_dataset/17_ListGPSCoordinates.csv", sep=",", header=T)
+
+# show data
+data %>% head(3) %>% kable() %>%
+ kable_styling(bootstrap_options = "striped", full_width = F)If your dataset is composed by a few data point only, you can just display them on a map. If you have specific information to display concerning these positions, use an interactive: more information are available when you click data points.
-If your dataset is composed by a few data point only, you can just +display them on a map. If you have specific information to display +concerning these positions, use an interactive: more information are +available when you click data points.
+A dot density map is used when the sample size of your dataset is high.
-# Get the world polygon
-world <- map_data("world")
-
-data %>%
- #head(1000) %>%
- ggplot( aes(x=homelon, y=homelat)) +
- geom_polygon(data = world, aes(x=long, y = lat, group = group), fill="grey", alpha=0.1) +
- geom_point(size=0.8, color="#69b3a2", alpha=0.5) +
- coord_equal() +
- theme_void() +
- theme(
- panel.spacing=unit(c(0,0,0,0), "null"),
- plot.margin=grid::unit(c(0,0,0,0), "cm"),
- ) +
- ggplot2::annotate("text", x = -150, y = -45, hjust = 0, size = 11, label = paste("Where surfers live."), color = "Black") +
- ggplot2::annotate("text", x = -150, y = -51, hjust = 0, size = 8, label = paste("data-to-viz.com | 200,000 #surf tweets recovered"), color = "black", alpha = 0.5) +
- xlim(-180,180) +
- ylim(-60,80) +
- scale_x_continuous(expand = c(0.006, 0.006)) +
- coord_equal() A dot density map is used when the sample size of your dataset is +high.
+# Get the world polygon
+world <- map_data("world")
+
+data %>%
+ #head(1000) %>%
+ ggplot( aes(x=homelon, y=homelat)) +
+ geom_polygon(data = world, aes(x=long, y = lat, group = group), fill="grey", alpha=0.1) +
+ geom_point(size=0.8, color="#69b3a2", alpha=0.5) +
+ coord_equal() +
+ theme_void() +
+ theme(
+ panel.spacing=unit(c(0,0,0,0), "null"),
+ plot.margin=grid::unit(c(0,0,0,0), "cm"),
+ ) +
+ ggplot2::annotate("text", x = -150, y = -45, hjust = 0, size = 11, label = paste("Where surfers live."), color = "Black") +
+ ggplot2::annotate("text", x = -150, y = -51, hjust = 0, size = 8, label = paste("data-to-viz.com | 200,000 #surf tweets recovered"), color = "black", alpha = 0.5) +
+ xlim(-180,180) +
+ ylim(-60,80) +
+ scale_x_continuous(expand = c(0.006, 0.006)) +
+ coord_equal() 
ggsave("IMG/Surfer_position.png", width = 36, height = 15.22, units = "in", dpi = 90)
Dotmaps give a good idea about where samples are distributed. However, once dots start to overlap, it gets impossible to distinguish how many of them are displayed on a certain zone. That’s where binning becomes an interesting option.
+Dotmaps give a good idea about where samples are distributed. +However, once dots start to overlap, it gets impossible to distinguish +how many of them are displayed on a certain zone. That’s where binning +becomes an interesting option.
To create a hexbin map, the territory is divided in many hexagones and the number of sample per hexagone is counted and represented by a color.
-data %>%
- filter(homecontinent=='Europe') %>%
- ggplot( aes(x=homelon, y=homelat)) +
- geom_hex(bins=65) +
- theme_void() +
- xlim(-30, 70) +
- ylim(24, 72) +
- scale_fill_viridis(
- trans = "log",
- name="Number of Tweet recorded in 8 months",
- guide = guide_legend( keyheight = unit(3, units = "mm"), keywidth=unit(12, units = "mm"), label.position = "bottom", title.position = 'top', nrow=1)
- ) +
- ggtitle( "Where people tweet about #Surf" ) +
- theme(
- legend.position = c(0.5, 0.09),
- text = element_text(color = "#22211d"),
- #plot.background = element_rect(fill = "#f5f5f2", color = NA),
- #panel.background = element_rect(fill = "#f5f5f2", color = NA),
- #legend.background = element_rect(fill = "#f5f5f2", color = NA),
- plot.title = element_text(size= 22, hjust=0.1, color = "#4e4d47", margin = margin(b = -0.1, t = 0.4, l = 2, unit = "cm")),
- )To create a hexbin map, the territory is divided in many hexagones +and the number of sample per hexagone is counted and represented by a +color.
+data %>%
+ filter(homecontinent=='Europe') %>%
+ ggplot( aes(x=homelon, y=homelat)) +
+ geom_hex(bins=65) +
+ theme_void() +
+ xlim(-30, 70) +
+ ylim(24, 72) +
+ scale_fill_viridis(
+ trans = "log",
+ name="Number of Tweet recorded in 8 months",
+ guide = guide_legend( keyheight = unit(3, units = "mm"), keywidth=unit(12, units = "mm"), label.position = "bottom", title.position = 'top', nrow=1)
+ ) +
+ ggtitle( "Where people tweet about #Surf" ) +
+ theme(
+ legend.position = c(0.5, 0.09),
+ text = element_text(color = "#22211d"),
+ #plot.background = element_rect(fill = "#f5f5f2", color = NA),
+ #panel.background = element_rect(fill = "#f5f5f2", color = NA),
+ #legend.background = element_rect(fill = "#f5f5f2", color = NA),
+ plot.title = element_text(size= 22, hjust=0.1, color = "#4e4d47", margin = margin(b = -0.1, t = 0.4, l = 2, unit = "cm")),
+ )
# Make the hexbin map with the geom_hex function
-ggplot(data, aes(x=homelon, y=homelat)) +
- geom_polygon(data = world, aes(x=long, y = lat, group = group), fill="grey", alpha=0.3) +
- geom_bin2d(bins=120) +
- theme_void() +
- scale_fill_viridis(
- trans = "log",
- name="Number of Tweets",
- guide = guide_legend( keyheight = unit(3, units = "mm"), keywidth=unit(12, units = "mm"), label.position = "bottom", title.position = 'top', nrow=1)
- ) +
- theme(
- legend.position = c(0.5, 0.09),
- ) +
- ylim(-70, 80)# Make the hexbin map with the geom_hex function
+ggplot(data, aes(x=homelon, y=homelat)) +
+ geom_polygon(data = world, aes(x=long, y = lat, group = group), fill="grey", alpha=0.3) +
+ geom_bin2d(bins=120) +
+ theme_void() +
+ scale_fill_viridis(
+ trans = "log",
+ name="Number of Tweets",
+ guide = guide_legend( keyheight = unit(3, units = "mm"), keywidth=unit(12, units = "mm"), label.position = "bottom", title.position = 'top', nrow=1)
+ ) +
+ theme(
+ legend.position = c(0.5, 0.09),
+ ) +
+ ylim(-70, 80)
It is also possible to divide your territory (the world here) in regions other than square or hexagones. This will result in a chloropleth map. Of course, you need the information of the exact shape of your regions.
+It is also possible to divide your territory (the world here) in +regions other than square or hexagones. This will result in a +chloropleth map. Of course, you need the information of the exact shape +of your regions.
You can learn more about each type of graphic presented in this story in the dedicated sections. Click the icon below:
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
- Any thoughts on this? Found any mistake? Have another way to show the data? Please drop me a word on Twitter or in the comment section below:
-+ You can learn more about each type of graphic presented in this + story in the dedicated sections. Click the icon below: +
+ +
+
+
+
+
+ + Data To Viz is a + comprehensive classification of chart types organized by data + input format. Get a high-resolution version of our decision tree + delivered to your inbox now! +
+
+ A work by Yan Holtz for data-to-viz.com
@@ -464,28 +539,28 @@
- Evolution of baby names in the US
-
-
- 
-
- 
-
- 
-
-
-
-
-
-
-
-
+ Evolution of baby names in the US
+
+
+
+
+
+
+
+
This document aims to describe a few dataviz that can be applied to a dataset containing an ordered numeric variable, a categoric variable and another numeric variable. An an example we will consider the evolution of baby name frequencies in the US between 1880 and 2015. This dataset is available through the babynames R library and a .csv version is available on github. It looks as follow:
This document aims to describe a few dataviz that can be
+applied to a dataset containing an ordered numeric variable, a categoric
+variable and another numeric variable. An an example we will consider
+the evolution of baby name frequencies in the US between 1880 and 2015.
+This dataset is available through the babynames R library and a
+.csv version is available on github.
+It looks as follow:
# Libraries
-library(tidyverse)
-library(hrbrthemes)
-library(kableExtra)
-options(knitr.table.format = "html")
-library(babynames)
-library(streamgraph)
-library(viridis)
-library(DT)
-library(plotly)
-
-# Load dataset from github
-data <- babynames %>%
- filter(name %in% c("Ashley", "Amanda", "Jessica", "Patricia", "Linda", "Deborah", "Dorothy", "Betty", "Helen")) %>%
- filter(sex=="F")
-
-# Show long format
-data %>%
- select(year, name, n) %>%
- head(5) %>%
- arrange(name) %>%
- kable() %>%
- kable_styling(bootstrap_options = "striped", full_width = F)# Libraries
+library(tidyverse)
+library(hrbrthemes)
+library(kableExtra)
+options(knitr.table.format = "html")
+library(babynames)
+library(streamgraph)
+library(viridis)
+library(DT)
+library(plotly)
+
+# Load dataset from github
+data <- babynames %>%
+ filter(name %in% c("Ashley", "Amanda", "Jessica", "Patricia", "Linda", "Deborah", "Dorothy", "Betty", "Helen")) %>%
+ filter(sex=="F")
+
+# Show long format
+data %>%
+ select(year, name, n) %>%
+ head(5) %>%
+ arrange(name) %>%
+ kable() %>%
+ kable_styling(bootstrap_options = "striped", full_width = F)#Histogram ***
+The most common way to represent a unique numeric variable is with a
histogram. Basically, the numeric variable is cut in several
bins: between 0 and 10 euros a night, between 10 and 20 and
@@ -374,7 +376,10 @@
Actually a few values are over represented in the dataset (like 58, 64,
69, 75, 80..). This is definitely a signal that you want to understand
when analysing your dataset.
#Density ***
+A variation of the histogram is the density plot, which is basically
a smoothed version of the histogram. It represents a
kernel density estimate of the variable. As seen for the
@@ -399,6 +404,7 @@
theme_ipsum()
diff --git a/story/OneNumOneCat.Rmd b/story/OneNumOneCat.Rmd
index e880a36..0c941cc 100644
--- a/story/OneNumOneCat.Rmd
+++ b/story/OneNumOneCat.Rmd
@@ -40,6 +40,7 @@ A clean version at `.csv` format is available on github, and the table beside sh
- Who sells more weapons?
-
-
- 
-
- 
-
- 
-
- 
-
-
-
-
-
-
-
+ Who sells more weapons?
+
+
+
+
+
+
+
+
+
+
This document gives a few suggestions to analyse a dataset composed by one numeric and one categoric variable.
The dataset used as an example provides the quantity of weapons exported by the top 50 largest exporters in 2017. This dataset has been downloaded on the Stockholm International Peace Research Institute website and more information about it are available here.
A clean version at .csv format is available on github, and the table beside shows how it looks like.
This document gives a few suggestions to analyse a
+dataset composed by one numeric and one categoric variable.
+
The dataset used as an example provides the quantity of weapons
+exported by the top 50 largest exporters in 2017. This dataset has been
+downloaded on the Stockholm
+International Peace Research Institute website and more information
+about it are available here.
+
A clean version at .csv format is available on github,
+and the table beside shows how it looks like.
# Libraries
-library(tidyverse)
-library(hrbrthemes)
-library(kableExtra)
-options(knitr.table.format = "html")
-
-# Load dataset from github
-data <- read.table("https://raw.githubusercontent.com/holtzy/data_to_viz/master/Example_dataset/7_OneCatOneNum.csv", header=TRUE, sep=",")
-
-# show data
-data %>% head(6) %>% kable() %>%
- kable_styling(bootstrap_options = "striped", full_width = F)# Libraries
+library(tidyverse)
+library(hrbrthemes)
+library(kableExtra)
+options(knitr.table.format = "html")
+
+# Load dataset from github
+data <- read.table("https://raw.githubusercontent.com/holtzy/data_to_viz/master/Example_dataset/7_OneCatOneNum.csv", header=TRUE, sep=",")
+
+# show data
+data %>% head(6) %>% kable() %>%
+ kable_styling(bootstrap_options = "striped", full_width = F)| -Improbable +We Doubt | 33 @@ -283,58 +340,58 @@ |
| -Almost Certainly +About Even | -98 +50 |
| -Likely +Almost No Chance | -60 +1 |
| -Almost Certainly +Highly Unlikely | -98 +5 |
| -Unlikely +Almost No Chance | -10 +2 |
| -Probably Not +Improbable | -25 +3 |
| -About Even +Chances Are Slight | -50 +40 |
| -Probably +We Doubt | -75 +10 |
Note that here the Y scale is the same for each group, unlike the precedent example on density plot.
