{chronicle}

A quick demo

You can build R Markdown reports through the add_* family of functions, layering one below the previous one.

library(chronicle)

demo_report <-
  add_text(text_title = "This is the output of a chronicle call",
           text = "Each element has been added through an add_* function.",
           title_level = 1) %>%
  add_table(table = head(iris),
            table_title = "A glimpse at the iris dataset",
            html_table_type = "kable",
            title_level = 1) %>%
  add_raincloud(dt = iris,
                value = "Sepal.Length",
                groups = "Species",
                raincloud_title = "Distribution of sepal length by species",
                title_level = 2) %>%
  add_scatterplot(dt = iris,
                  x = "Petal.Width",
                  y = "Petal.Length",
                  groups = "Species",
                  scatterplot_title = "Comparison of petal width and length",
                  title_level = 2)

render_report(report = demo_report,
              output_format = "rmdformats",
              filename = "quick_demo",
              title = "A quick chronicle demo",
              author = "You did this!",
              keep_rmd = TRUE)

## Warning: MathJax doesn't work with self_contained when not using the rmarkdown
## "default" template.

## [1] TRUE

## [1] TRUE

You can see the output of this call, and a full showcase of all the elements supported by chronicle.

What happens behind these calls is that chronicle writes an R Markdown for you! you can see the report we’ve built by calling it through cat()

demo_report

## 
## 
## # This is the output of a chronicle call
## 
## Each element has been added through an add_* function.
## 
## # A glimpse at the iris dataset
## ```{r, echo=FALSE, message=FALSE, warning=FALSE}
## knitr::kable(head(iris))
## ```
## 
## ## Distribution of sepal length by species
## ```{r, echo=FALSE, message=FALSE, warning=FALSE, fig.width=params$figure_width, fig.height=params$figure_height}
## chronicle::make_raincloud(dt = iris,
##                           value = 'Sepal.Length',
##                           groups = 'Species',
##                           adjust = 0.5,
##                           include_boxplot = TRUE,
##                           include_mean = FALSE,
##                           include_median = TRUE,
##                           force_all_jitter_obs = FALSE,
##                           ggtheme = 'minimal',
##                           plot_palette = params$plot_palette,
##                           plot_palette_generator = params$plot_palette_generator,
##                           static = params$set_static)
## ```
## 
## ## Comparison of petal width and length
## ```{r, echo=FALSE, message=FALSE, warning=FALSE, fig.width=params$figure_width, fig.height=params$figure_height}
## chronicle::make_scatterplot(dt = iris,
##                             x = 'Petal.Width',
##                             y = 'Petal.Length',
##                             groups = 'Species',
##                             plot_palette = params$plot_palette,
##                             plot_palette_generator = params$plot_palette_generator,
##                             static = params$set_static)
## ```

The make_* family of functions

Every plot added with an add_* function will be built through its correpsonding make_* function. These functions take care of the heavy lifting, avoiding the cumbersome (albeit powerful) sintax of ggplot, plotly and other html widgets. The parameters of the make_functions are simple and intuitive specifications on how to make each plot, and they can be called independently and used in any instance where a ggplot or an html widget would fit.

make_barplot(dt = ggplot2::mpg,
             value = 'cty',
             bars = 'manufacturer',
             break_bars_by = 'drv',
             horizontal = TRUE,
             sort_by_value = TRUE,
             static = TRUE)

chronicle bar plot

make_raincloud(dt = iris,
             value = 'Sepal.Length',
             groups = 'Species')

chronicle rain cloud plot

Rendering chronicle reports

Once the structure of the report has been defined, the rendering process is done by render_report(). This uses rmarkdonw::render() as a backend for rendering the report, which gives chronicle the capability to render the reports with full visibility to all objects in the global environment. This gives chronicle two of its main strengths:

1. You don’t need to include nor run all your data processing code again for a new report output. This means you can build several report recipes for different audiences out of the same data processing, with each one having their own report recipe.

2. It can render several output formats in a single call. For instance, it is possible to render the same content as ioslides for a presentation, as tufte_html for handouts and as rmdformats for a site upload.

Take our quick demo as an example, to render this as the three outputs mentioned previously, you only need to add that vector to the output_format parameter of render_report()

render_report(report = demo_report,
              output_format = c("ioslides", "tufte_html", "rmdformats"),
              filename = "quick_demo",
              title = "A quick chronicle demo",
              author = "You did this!",
              keep_rmd = TRUE)

The report_columns() function

chronicle also includes a function called report_columns(), that will create an entire chronicle report for a single dataset. It includes a comprehensive summary of the data through the skimr::skim() function, along with one plot for each column present in the data: bar plots for categorical variables and rain cloud plots for numerical variables. This gives you an immediate view of a dataset with a single line of code!

report_columns(dt = palmerpenguins::penguins,
               by_column = 'species')

And you can see the output of this call

Supported formats

As of version 0.2.5, chronicle can output both static and dynamic outputs. Dynamic outputs refer to R Markdown formats that support html widgets, hence the elements added will be dynamic plots (plotly, dygraph, DT). For static outputs, these will roll back to ggplot and static table prints.

Supported report elements

I highly encourage you to review the enitre showcase, words are not as adequate to describe each element. But for a quick glance, as of version 0.2.5 chronicle supports:

• Bar plots
• Box plots
• Code (optionally evaluated)
• Density plots
• Dygraphs
• Histograms
• Line plots
• Rain cloud plots
• Scatter plots
• Tables
• Texts
• Titles
• Violin plots