In graggsd/tidytableR: Builds Tables from Tidy Datasets

Introduction

tidy_htmlTable acts as a wrapper function for the htmlTable function allowing columns to be mapped from the input data to specific htmlTable parameters in a manner similar to ggplot2.

tidy_df acts in a similar manner, except that it outputs a data.frame to make saving its output into tabular files easier.

Example Use

Prepare Data

We'll begin by turning the mtcars data into a tidy dataset. The gather function is called to collect 3 performance metrics into a pair of key and value columns.

library(tidytableR)
library(magrittr)
library(dplyr)
library(tidyr)
library(tibble)

td <- mtcars %>%
    rownames_to_column %>%
    select(rowname, cyl, gear, hp, mpg, qsec) %>%
    gather(per_metric, value, hp, mpg, qsec)

Now we will compute 4 summary statistics for each of the 3 performance metrics. This will be further grouped by number of cylinders and gears.

tidy_summary <- td %>%
    group_by(cyl, gear, per_metric) %>% 
    summarise(Mean = round(mean(value), 1),
              SD = round(sd(value), 1),
              Min = round(min(value), 1),
              Max = round(max(value), 1)) %>%
    gather(summary_stat, value, Mean, SD, Min, Max) %>% 
    ungroup %>% 
    mutate(gear = paste(gear, "Gears"),
           cyl = paste(cyl, "Cylinders"))

Though in this format, tidy_summary is easy to pass onto functions within ggplot2, with 86 rows, it is not conducive to presentating as a table. Though tidyr may be used to create a wide format, this will take some work, and there will still be additional steps if one wishes to present the data using the htmlTable package. The tidytableR package was created to shortcut these steps.

`tidy_htmlTable`

Columns from tidy_summary may be mapped directly to arguments within the htmlTable function similar to the way that columns are mapped to elements of a plot within ggplot2. That said, to use functions within this package appropriately, it is important to be familiar with the htmlTable package.

Note that, at minimum, the value, rnames, and header arguments must be specified.

Also, the combination of rnames and header (additionally, rgroup, tspanner, cgroup1, and cgroup2, if they are specified by the user) must form a unique key for each row. This will most often occur naturally if the the input data has been created dplyr::summarise or tidyr::gather, which are the primary use cases for these functions.

Example 1

tidy_summary  %>% 
    tidy_htmlTable(header = "gear",
                   cgroup1 = "cyl",
                   cell_value = "value", 
                   rnames = "summary_stat",
                   rgroup = "per_metric")

Example 2

tidy_summary  %>% 
    tidy_htmlTable(header = "summary_stat",
                 cgroup1 = "per_metric",
                 cell_value = "value", 
                 rnames = "gear",
                 rgroup = "cyl")

`tidy_df`

Often, a tabular data file will be desired for sharing or manual editing. tidy_df has the same functionality as tidy_htmlTable, but returns a data.frame instead of html code.

Without collapsing

By default, the output of tidy_df will include extra columns on the left margin of the output data.frame to designate values for tspanner, rgroup, and rnames (ordered from outside to inside respectively), and the top margin to designate cgroup2, cgroup1, and header values (ordered from outside to inside respectively).

tidy_summary  %>%
    # So that output may be displayed appropriately
    filter(cyl != "6 Cylinders") %>% 
    tidy_df(header = "gear",
            cgroup1 = "cyl",
            cell_value = "value", 
            rnames = "summary_stat",
            rgroup = "per_metric")

With collapsing

As an alternative, a character string may be specified in the collapse argument, which will result in tspanner, rgroup, and rnames being combined into a single term collapse and included as a single column on the far left margin of the table, and cgroup2, cgroup1, and header receiving the same treatment, but being used to name columns of the output data.frame.

tidy_summary  %>%
    # So that output may be displayed appropriately
    filter(cyl != "6 Cylinders",
           cyl != "8 Cylinders", 
           gear != "5 Gears") %>% 
    tidy_df(header = "gear",
            cgroup1 = "cyl",
            cell_value = "value", 
            rnames = "summary_stat",
            rgroup = "per_metric",
            collapse = ".")