Business and Biodiversity - Ch 2 - Snowball - Pilot 2

Setup

Code

#|

suppressPackageStartupMessages({
    library(tibble)
    library(bibtex)
    library(openalexR)
    library(IPBES.R)
    library(dplyr)
    library(knitr)
    library(networkD3)
    library(ggplot2)
    library(tidygraph)
    library(readxl)
    library(grDevices)
})

file.create("figures_comp", showWarnings = FALSE)

[1] FALSE

Back to Readme

Read snowball results from Ch2 and Ch 4 2

Code

selected_papers <- readxl::read_excel(
    file.path(".", "data_ch2", "selected_papers.xlsx"),
    skip = 1
)$id

New names:
• `` -> `...3`
• `` -> `...4`
• `` -> `...10`
• `` -> `...11`

Code

kp_ids <- readRDS(file.path(".", "data_ch2", "key_works.rds"))$id |>
    gsub(pattern = "https://openalex.org/", replacement = "")

snowball_ch2 <- readRDS(
    file.path(".", "data_ch2", "snowball.rds")
)
snowball_ch2$nodes <- snowball_ch2$nodes |>
    dplyr::filter(id %in% selected_papers) |>
    dplyr::mutate(
        cited_by_count_by_year = cited_by_count / (2024 - publication_year),
        chapter = "Chapter 2",
        key_paper = id %in% kp_ids,
        paper_type = ifelse(
            key_paper,
            "Ch 2 KP",
            "Ch 2"
        )
    )
snowball_ch2$nodes$auth_abbr <- IPBES.R::abbreviate_authors(snowball_ch2$nodes)
snowball_ch2$edges <- snowball_ch2$edges |>
    dplyr::filter((from %in% selected_papers) & (to %in% selected_papers))

# flat_snow_ch2 <- snowball2df(snowball_ch2) |>
#     tibble::as_tibble()

rm(selected_papers, kp_ids)

kp_ids <- readRDS(file.path(".", "data", "key_works.rds"))$id |>
    gsub(pattern = "https://openalex.org/", replacement = "")
snowball_ch4_2 <- readRDS(
    file.path(".", "data", "snowball.rds")
)
snowball_ch4_2$nodes <- snowball_ch4_2$nodes |>
    dplyr::mutate(
        cited_by_count_by_year = cited_by_count / (2024 - publication_year),
        chapter = "Chapter 4",
        key_paper = id %in% kp_ids,
        paper_type = ifelse(
            key_paper,
            "Ch 4 2 KP",
            "Ch 4 2"
        )
    )
snowball_ch4_2$nodes$auth_abbr <- IPBES.R::abbreviate_authors(snowball_ch4_2$nodes)
# flat_snow_c4_2 <- snowball2df(snowball_ch4_2) |>
#     tibble::as_tibble()

rm(kp_ids)

Merge two snowballs into one

Code

snowball <- snowball_ch2

snowball$nodes <- rbind(snowball$nodes, snowball_ch4_2$nodes) |>
    dplyr::group_by(id, .drop = FALSE) |>
    mutate(chapter = paste(chapter, collapse = " & ")) |>
    mutate(paper_type = paste(paper_type, collapse = " & ")) |>
    ungroup() |>
    distinct(id, .keep_all = TRUE)
snowball$edges <- rbind(snowball$edges, snowball_ch4_2$edges) |>
    dplyr::distinct()

connected <- snowball$edges |>
    unlist() |>
    unique()

snowball_connected <- snowball
snowball_connected$nodes <- snowball_connected$nodes |>
    dplyr::filter(id %in% connected)

rm(connected)

Static Plots

All included Papers

All references. The points are only in Chapter 2 and their kay papers have been removed from the snowball corpus.

Code

p <- snowball |>
    tidygraph::as_tbl_graph() |>
    ggraph::ggraph(graph = , layout = "stress") +
    ggraph::geom_edge_link(
        aes(
            alpha = ggplot2::after_stat(index),
            # edge_width = ggplot2::after_stat(index)
        ),
        show.legend = FALSE
    ) +
    ggraph::geom_node_point(
        ggplot2::aes(
            size = cited_by_count_by_year,
            col = paper_type
        )
    ) +
    ggraph::geom_node_label(
        ggplot2::aes(
            filter = key_paper,
            label = auth_abbr
        ),
        nudge_y = 0.2,
        size = 3
    ) +
    ggraph::scale_edge_width(
        range = c(0.1, 1.5),
        guide = "none"
    ) +
    ggplot2::scale_size(
        range = c(3, 10),
        name = "Cited by average count per year"
    ) +
    ggplot2::scale_colour_manual(
        values = c("red", "blue", "green", "yellow", "purple", "orange", "pink", "brown"),
        na.value = "grey",
        name = "Key Paper",
        guide = guide_legend()
    ) +
    ggraph::theme_graph() +
    ggplot2::theme(
        plot.background = element_rect(fill = "transparent", colour = NA),
        panel.background = element_rect(fill = "transparent", colour = NA),
        legend.position = "right"
    ) +
    ggplot2::guides(fill = "none") +
    ggplot2::ggtitle(paste0(" Cited by average count per year"))

ggplot2::ggsave(
    file.path("figures_comp", "all.pdf"),
    plot = p,
    device = grDevices::cairo_pdf,
    width = 20,
    height = 15
)

Warning: Using the `size` aesthetic in this geom was deprecated in ggplot2 3.4.0.
ℹ Please use `linewidth` in the `default_aes` field and elsewhere instead.

Code

ggplot2::ggsave(
    file.path("figures_comp", "all.png"),
    plot = p,
    width = 20,
    height = 15,
    bg = "white"
)

A highres version can be downloaded here.

Connected Papers

A subset excluding the references which are not part of the network but isolated points.

Code

p <- snowball_connected |>
    tidygraph::as_tbl_graph() |>
    ggraph::ggraph(graph = , layout = "stress") +
    ggraph::geom_edge_link(
        aes(
            alpha = ggplot2::after_stat(index),
            # edge_width = ggplot2::after_stat(index)
        ),
        show.legend = FALSE
    ) +
    ggraph::geom_node_point(
        ggplot2::aes(
            size = cited_by_count_by_year,
            col = paper_type
        )
    ) +
    ggraph::geom_node_label(
        ggplot2::aes(
            filter = key_paper,
            label = auth_abbr
        ),
        nudge_y = 0.2,
        size = 3
    ) +
    ggraph::scale_edge_width(
        range = c(0.1, 1.5),
        guide = "none"
    ) +
    ggplot2::scale_size(
        range = c(3, 10),
        name = "Cited by average count per year"
    ) +
    ggplot2::scale_colour_manual(
        values = c("red", "blue", "green", "yellow", "purple", "orange", "pink", "brown"),
        na.value = "grey",
        name = "Key Paper",
        guide = guide_legend()
    ) +
    ggraph::theme_graph() +
    ggplot2::theme(
        plot.background = element_rect(fill = "transparent", colour = NA),
        panel.background = element_rect(fill = "transparent", colour = NA),
        legend.position = "right"
    ) +
    ggplot2::guides(fill = "none") +
    ggplot2::ggtitle(paste0(" Cited by average count per year"))

ggplot2::ggsave(
    file.path("figures_comp", "connected.pdf"),
    plot = p,
    device = grDevices::cairo_pdf,
    width = 20,
    height = 15
)
ggplot2::ggsave(
    file.path("figures_comp", "connected.png"),
    plot = p,
    width = 20,
    height = 15,
    bg = "white"
)

A highres version can be downloaded here.

Dynamic plot

Connecter Papers Coloured by Chapter

And some dynamic plot of the connected dataset

Code

#|
networkData <- data.frame(
    src = snowball_connected$edges$from,
    target = snowball_connected$edges$to,
    stringsAsFactors = FALSE
)

nodes <- data.frame(
    name = snowball_connected$nodes$id,
    author = IPBES.R::abbreviate_authors(snowball_connected$nodes),
    doi = snowball_connected$nodes$doi,
    nodesize = snowball_connected$nodes$cited_by_count / (2024 - snowball_connected$nodes$publication_year) * 0.5,
    paper_type = snowball_connected$nodes$paper_type,
    chapter = snowball_connected$nodes$chapter,
    stringsAsFactors = FALSE
)
nodes$id <- 0:(nrow(nodes) - 1)

# create a data frame of the edges that uses id 0:9 instead of their names
edges <- networkData |>
    left_join(nodes, by = c("src" = "name")) |>
    select(-src, -author) |>
    rename(source = id) |>
    left_join(nodes, by = c("target" = "name")) |>
    select(-target, -author) |>
    rename(target = id) |>
    mutate(width = 1)

# make a grouping variable that will match to colours
nodes$group <- nodes$chapter

nodes$oa_id <- nodes$name
nodes$name <- nodes$author

# control colours with a JS ordinal scale
ColourScale <- 'd3.scaleOrdinal()
                        .domain([
                            "Chapter 2",
                            "Chapter 4",
                            "Chapter 2 & Chapter 4",
                        ])
                        .range(["blue", "red", "green"]);'

openDOI <- "window.open(d.doi)"

MyClickScript <- 'alert("You clicked " + d.name + " with the doi " +
             d.doi +  " of your original R data frame");'

nwg_ch <- forceNetwork(
    Links = edges,
    Nodes = nodes,
    Source = "source",
    Target = "target",
    NodeID = "name",
    Nodesize = "nodesize",
    Group = "chapter",
    Value = "width",
    opacity = 0.9,
    zoom = TRUE,
    colourScale = JS(ColourScale),
    fontSize = 40,
    legend = TRUE,
    clickAction = openDOI
)

nwg_ch$x$nodes$doi <- nodes$doi

networkD3::saveNetwork(
    nwg_ch,
    file = file.path("figures_comp", "connected_chapter.html"),
    selfcontained = TRUE
)

unlink(
    list.dirs(path = "figures_comp", recursive = FALSE),
    recursive = TRUE
)

nwg_ch

You can download the interactive graph here (right mouse click).

Connecter Papers Coloured by Chapter and Keypaper

And some dynamic plot of the connected dataset

Code

#|

networkData <- data.frame(
    src = snowball_connected$edges$from,
    target = snowball_connected$edges$to,
    stringsAsFactors = FALSE
)

nodes <- data.frame(
    name = snowball_connected$nodes$id,
    author = IPBES.R::abbreviate_authors(snowball_connected$nodes),
    doi = snowball_connected$nodes$doi,
    nodesize = snowball_connected$nodes$cited_by_count / (2024 - snowball_connected$nodes$publication_year) * 0.5,
    paper_type = snowball_connected$nodes$paper_type,
    stringsAsFactors = FALSE
)
nodes$id <- 0:(nrow(nodes) - 1)

# create a data frame of the edges that uses id 0:9 instead of their names
edges <- networkData |>
    left_join(nodes, by = c("src" = "name")) |>
    select(-src, -author) |>
    rename(source = id) |>
    left_join(nodes, by = c("target" = "name")) |>
    select(-target, -author) |>
    rename(target = id) |>
    mutate(width = 1)

# make a grouping variable that will match to colours
nodes$group <- nodes$paper_type

nodes$oa_id <- nodes$name
nodes$name <- nodes$author

# control colours with a JS ordinal scale
ColourScale <- 'd3.scaleOrdinal()
                        .domain([
                            "Ch 2 KP & Ch 4 2 KP",
                            "Ch 2 KP", "Ch 2 KP & Ch 4 2", "Ch 2",
                            "Ch 4 2 KP", "Ch 2 KP & Ch 4 2 KP", "Ch 4 2",
                        ])
                        .range(["red", "blue", "green", "yellow", "purple", "orange", "pink", "brown"]);'

openDOI <- "window.open(d.doi)"

MyClickScript <- 'alert("You clicked " + d.name + " with the doi " +
             d.doi +  " of your original R data frame");'

nwg <- forceNetwork(
    Links = edges,
    Nodes = nodes,
    Source = "source",
    Target = "target",
    NodeID = "name",
    Nodesize = "nodesize",
    Group = "paper_type",
    Value = "width",
    opacity = 0.9,
    zoom = TRUE,
    colourScale = JS(ColourScale),
    fontSize = 40,
    legend = TRUE,
    clickAction = openDOI
)

nwg$x$nodes$doi <- nodes$doi

networkD3::saveNetwork(
    nwg,
    file = file.path("figures_comp", "connected.html"),
    selfcontained = TRUE
)

unlink(
    list.dirs(path = "figures_comp", recursive = FALSE),
    recursive = TRUE
)

nwg

You can download the interactive graph here (right mouse click).

Tables

Table of papers selected by Chapter 2 and all Chapter 4

Code

library(DT)

Attaching package: 'DT'

The following object is masked from 'package:networkD3':

    JS

Code

snowball$nodes |>
    dplyr::select(
        chapter,
        paper_type,
        id,
        auth_abbr,
        publication_year,
        display_name,
        doi,
        cited_by_count,
        cited_by_count_by_year
    ) |>
    mutate(doi = sprintf('<a href="%s" target="_blank">%s</a>', doi, doi)) |>
    datatable(
        extensions = c(
            "Buttons",
            "FixedColumns",
            "Scroller"
        ),
        options = list(
            dom = "Bfrtip",
            buttons = list(
                "copy", 
                "csv", 
                "excel", 
                list(
                    extend = "pdf",
                    orientation = "landscape",
                    customize = JS(
                        "function(doc) {",
                         "  doc.defaultStyle.fontSize = 5;",  # Change the font siz
                         "}"
                    )
                ), 
                "print"
            ),
            scroller = TRUE,
            scrollY = JS("window.innerHeight * 0.7 + 'px'"),
            scrollX = TRUE,
            fixedColumns = list(leftColumns = 2)
        ),
        escape = FALSE
    )