fplot is an R package made for the easy plotting of distributions.
The problem when it comes to distributions is that we often face a variety of data, each requiring a different type of layout: we don’t represent the distribution of categorical variables similarly to continuous variables, and regarding the latter heavily skewed data may require a special treatment. The aim of fplot is to automatically make choices for the user in order to always display meaningful (and hopefully beautiful!) graphs.
This brief description shows a gallery of fplot possibilities. For a more detailed introduction, please see the walk-through.
The data sets used are us_pub_econ (from fplot) relating to publications in economics from US institutions, trade (from package fixest) relating to trade value between countries for some products in the European Union, and the iris data set (from base R).
One aim of fplotis to easily export graphs, it therefore includes a tool to relabel the variable names globally (once and for all). The following line is run before plotting the graphs:
setFplot_dict(c(Origin = "Exporting Country", Destination = "Importing Country", Euros = "Exports Value in €",
86 | jnl_top_25p = "Pub. in Top 25% journal", jnl_top_5p = "Publications in Top 5% journal",
87 | journal = "Journal", institution = "U.S. Institution", Petal.Length = "Petal Length"))The code to run the plots are in the titles of the graphs. You’ll notice that the same command line can result in a set of different graphs.
89 |
fit.off.RdThis function is deprecated: Please use the functions export_graph_start()
61 | and export_graph_end() instead.
fit.off()This function does not return anything in R. It closes the connection between the 73 | R graphics engine and a file that has been defined via one of the functions: 74 | pdf_fitpng_fit
75 |To be used in combination with pdf_fit or png_fit
79 | when exporting images. It performs exactly the same thing as dev.off() but additionaly
80 | shows the resulting graph in the viewer pane provided you're using RStudio.
To view the results of PDF exports, the function pdf_convert from package pdftools
82 | is used to convert the PDF files into images -- so you need to have installed
83 | pdftools to make it work.
In PDFs, only the first page will be viewed.
85 |The tool to set the page size and the exporting defaults: setFplot_page.
89 | Exporting functions pdf_fit, png_fit,
90 | jpeg_fit.
The functions export_graph_start() and export_graph_end() provide similar features.
101 | # Exportation example
102 | # The functions pdf_fit, png_fit, etc, guarantee the right
103 | # point size of the texts present in the graph.
104 | # But you must give the exact size the graph will take in your final document.
105 | # => first use the function setFplot_page, default is:
106 | # setFplot_page(page = "us", margins = "normal")
107 | # By default the graph takes 100% of the text width
108 |
109 | data(us_pub_econ)
110 |
111 | tmpFile = file.path(tempdir(), "DISTR -- institutions.png")
112 |
113 | png_fit(tmpFile)
114 | plot_distr(~institution, us_pub_econ)
115 | fit.off()
116 |
117 | # What's the consequence of increasing the point size of the text?
118 | png_fit(tmpFile, pt = 15)
119 | plot_distr(~institution, us_pub_econ)
120 | fit.off()
121 |
122 |
123 | fplot-package.Rdfplot provides automatic plotting of common graphs (distributions, lines, bar plots and boxplots). The syntax uses formulas, allowing aggregate/conditional/weighted graphs with minimum efforts. The many arguments are automatically adjusted to the data in order to provide the nicest and most meaningful graphs.
60 |The core functions is plot_distr to draw distributions.
66 | Two other graphical functions are provided for convenience: plot_lines to represent the (usually temporal) evolution of some variables, and plot_box to easily represent conditional boxplots.
It also integrates tools to easily export graphs: pdf_fit and png_fit. In these functions, instead of providing the size of the graphics, you instead give the point size that the text should have in the final document--because an exported graph usually ends up in a document. You can set the size of your document with the function setFplot_page. If you use the function fit.off to close the connection, you will also see how the export looks like in the Viewer pane.
57 | All functions58 | 59 | |
60 | |
|---|---|
| 61 | 62 | | 63 |Graph export with garanteed text size |
64 |
| 65 | 66 | | 67 |Closes the current plotting device and shows the result in the viewer |
68 |
| 69 | 70 | | 71 |Aggregate/conditional graphs and automatic layout using formulas |
72 |
| 73 | 74 | | 75 |PDF export with guaranteed text size |
76 |
| 77 | 78 | | 79 |Boxplots with possibly moderators |
80 |
| 81 | 82 | | 83 |Plot distributions, possibly conditional |
84 |
| 85 | 86 | | 87 |Display means conditionnally on some other values |
88 |
| 89 | 90 | | 91 |PNG export with guaranteed text size |
92 |
| 93 | 94 | | 95 |Sets/gets the dictionary used in |
96 |
| 97 | 98 | | 99 |Sets the defaults of plot_distr |
100 |
| 101 | 102 | | 103 |Sets the target page size for figure exporting |
104 |
| 105 | 106 | | 107 |Publication data sample |
108 |
fplotsetFplot_dict.RdSets/gets the default dictionary used to rename the axes/moderator variables
65 | in the functions of the package fplot. The dictionaries are used to relabel
66 | variables (usually towards a fancier, more explicit formatting) that can be useful
67 | not to explicitly use the arguments xlab/ylab when exporting graphs. By setting
68 | the dictionary with setFplot_dict, you can avoid providing the argument
69 | dict in fplot functions.
setFplot_dict(dict)
74 |
75 | getFplot_dictAn object of class function of length 1.
A named character vector. E.g. to change my variable named "us_md"
86 | and "state" to (resp.) "$ miilion" and "U.S. state", then use
87 | dict = c(us_md="$ million", state = "U.S. state").
The function setFplot_dict() does not return anything, it only sets an option after checking
95 | the format of the arguments.
The function getFplot_dict() returns a named vector representing the
99 | dictionary set in setFplot_dict().
This function stores a named vector in the option "fplot_dict".
104 | The dictionary is automatically accessed by all fplot functions.
114 | data(airquality)
115 | setFplot_dict(c(Ozone = "Ozone (ppb)"))
116 | plot_distr(Ozone ~ Month, airquality, weight.fun = mean)
117 | #> NOTE: 37 observations with NAs (weight: 37)
118 | 
119 |
120 | us_pub_econ.RdThis data reports the publications of U.S. institutions in the field of economics between 1985 and 1990.
60 |data(us_pub_econ)us_pub_econ is a data table with 30,756 observations and 6 variables.
paper_id: Numeric identifier of the publication.
year: Year of publication.
institution: Institution of the authors of the publication.
journal: Journal/conference name.
jnl_top_25p: 0/1 variable of whether the journal belongs to the top 25% in terms of average cites.
jnl_top_5p: 0/1 variable of whether the journal belongs to the top 5% in terms of average cites.
The source is Microsoft Academic Graph (see reference).
79 |Arnab Sinha, Zhihong Shen, Yang Song, Hao Ma, Darrin Eide, Bo-June (Paul) Hsu, and Kuansan Wang. 2015. An Overview of Microsoft Academic Service (MAS) and Applications. In Proceedings of the 24th International Conference on World Wide Web (WWW '15 Companion). ACM, New York, NY, USA, 243-246.
83 |