Using plot_articles()

library(MDPIexploreR)

Overview

The function plot_articles() uses the data frame resulting from using article_info() to plot predefined graph types to summarize different The function plot_articles() uses the data frame resulting from article_info() to create predefined graph types that summarize various aspects of the data. In addition to the output of article_info(), plot_articles() requires:

  1. A string specifying the journal name for labeling the plot.
  2. A type argument to select the desired graph type.

Currently, five graph types are available. Below, we demonstrate these options using the Agriculture dataset obtained via article_info().

1. type = “summary”

This type option generates a dot graph, where each dot is a publication, against the year it was published.

plot_articles(agriculture,"Agriculture",type="summary")

2. type = “tat”

This type option shows the monthly changes in turnaround times (times between submission and article acceptance)

plot_articles(agriculture,"Agriculture",type="tat")
#> Scale for x is already present.
#> Adding another scale for x, which will replace the existing scale.

3. type = “issues”

This type option shows the yearly evolution on proportions of special issues, sections, topics and normal articles.

plot_articles(agriculture,"Agriculture",type="issues")

4. type = “review”

Coming soon

5. type = “type”

This type option shows the yearly evolution of types of articles published (Editorials, reviews,etc).

plot_articles(agriculture,"Agriculture",type="issues")

The function plot_articles() provides predefined graph types designed to summarize key aspects of the data quickly and effectively. However, users are encouraged to explore the data frame produced by article_info() to create custom visualizations tailored to their specific needs. The flexibility of the output makes it easy to integrate with popular R visualization libraries such as ggplot2 or plotly.

If you develop unique or insightful graphs that could benefit other users, consider submitting your code or ideas to contribute to enhancing this package.