In rOpenGov/bibliographica: Bibliographic Data Analysis

Publishers

• r length(unique(df$publisher)) unique publishers

• r sum(!is.na(df$publisher)) documents have unambiguous publisher information (r round(100*mean(!is.na(df$publisher)), 1)%). This includes documents identified as self-published; the author name is used as the publisher in those cases (if known).

• r length(which(df$self_published)) documents are identified as self-published (r round(100*(length(which(df$self_published))/nrow(df)), 1)%).

• Discarded publisher entries

• Conversions from original to final names (only non-trivial conversions shown)

The r ntop most common publishers are shown with the number of documents.

p <- top_plot(df, "publisher", ntop)
p <- p + ggtitle(paste("Top publishers"))
p <- p + scale_y_log10()
p <- p + ylab("Documents")
print(p)

Publication timeline for top publishers

Title count

# For some reason the ggplot never completes if top n > 2 ???
theme_set(theme_bw(20))
field <- "publisher"
tops <- names(top(df, field, n = 2))
dfs <- df %>% select(field, publication_decade) 
dfs$field <- as.character(dfs[[field]]); dfs[[field]] <- NULL
dfs <- dfs %>% filter(field %in% tops) %>%
               filter(!is.na(field) & !is.na(publication_decade)) %>%
           count(field, publication_decade)
dfs$field <- as.factor(dfs$field)
p <- ggplot(dfs, aes(x = publication_decade, y = n, fill = field)) +
       geom_bar(stat = "identity", position = "stack", color = "black") +       
       labs(x = "Publication Decade", y = "Title Count",
            title = "Title count timeline for the top publishers") +
       scale_fill_grey() +
       guides(fill = guide_legend("Publisher")) 
# print(p)

Title count versus paper consumption (top publishers):

#res <- compare_title_paper(df, "publisher", selected = tops)
tops <- names(top(df, field = "publisher", n = 10))
res <- compare_title_paper(df, "publisher")
print(res$plot)  
kable(subset(res$table, publisher %in% tops), digit = 2)

Corporates

Summaries of the corporate field.

• r length(unique(df$corporate)) unique corporates

• r sum(!is.na(df$corporate)) documents have unambiguous corporate information (r round(100*mean(!is.na(df$corporate)), 1)%).

• Discarded corporate entries

• Conversions from original to final names (only non-trivial conversions shown)

The r ntop most common corporates are shown with the number of documents.

p <- top_plot(df, "corporate", ntop)
p <- p + ggtitle(paste("Top corporates"))
p <- p + scale_y_log10()
p <- p + ylab("Documents")
print(p)

rOpenGov/bibliographica documentation built on April 10, 2022, 8:51 p.m.