Great Looking Tables: gt (v0.2)

栏目: IT技术 · 发布时间: 4年前

内容简介:We are extremely excited to have our first release of theIf you ever need to make beautiful customized

We are extremely excited to have our first release of the gt package available in CRAN! The name gt is short for “grammar of tables” and the goal of gt is similar to that of ggplot2 , serving to not just to make it easy to make specific tables, but to describe a set of underlying components that can be recombined in different ways to solve different problems.

If you ever need to make beautiful customized display tables, I think you’ll find gt is up to the task. You can install gt 0.2 from CRAN with:

install.packages("gt")

For an initial release, it’s pretty big! There are so many ways to structure a table, apply formatting and annotations, and style it just the way you want. Currently gt renders tables to the HTML output format (and has the ability to export to image files). We plan to also support the LaTeX and RTF output formats in the near future.

The website for the gt package has walkthrough articles for getting started and a function reference section with plenty of examples and images to show you how the table output is meant to appear.

Let’s get acquainted with our model of a table

We decided to formalize the parts of a table—and give them names—so that we have some language to act on. The larger components of a table (roughly from top to bottom) include the table header , the column labels , the stub and stub head , the table body , and the table footer . Within each of these components, there may be subcomponents (e.g., the table header contains a title and subtitle , the table body contains individual cells , etc.). Understanding how the parts fit together will make more sense with this diagram:

Great Looking Tables: gt (v0.2)

Learning new vocabulary is definitely a pain, but we believe it’s worthwhile. Like ggplot2 , the new words take some getting used to, but we believe learning them will improve your ability to analyze and understand existing tables, and then successfully recreate them in gt .

Examples with the exibble dataset

The exibble dataset is included in gt and its raison d’etre is to be a small dataset (8 rows and 9 columns) with different column types for experimenting with formatting. It fits easily on a single screen when printed as a tibble and rendered as a gt table, making it easy to see the results of our gt experimentation.

exibble

## # A tibble: 8 x 9
##           num char      fctr  date     time  datetime       currency row   group
##         <dbl> <chr>     <fct> <chr>    <chr> <chr>             <dbl> <chr> <chr>
## 1       0.111 apricot   one   2015-01… 13:35 2018-01-01 02…    50.0  row_1 grp_a
## 2       2.22  banana    two   2015-02… 14:40 2018-02-02 14…    18.0  row_2 grp_a
## 3      33.3   coconut   three 2015-03… 15:45 2018-03-03 03…     1.39 row_3 grp_a
## 4     444.    durian    four  2015-04… 16:50 2018-04-04 15… 65100    row_4 grp_a
## 5    5550     <NA>      five  2015-05… 17:55 2018-05-05 04…  1326.   row_5 grp_b
## 6      NA     fig       six   2015-06… <NA>  2018-06-06 16…    13.3  row_6 grp_b
## 7  777000     grapefru… seven <NA>     19:10 2018-07-07 05…    NA    row_7 grp_b
## 8 8880000     honeydew  eight 2015-08… 20:20 <NA>               0.44 row_8 grp_b

A simple table

Let’s use that dataset to make the ‘Hello, World!’ of gt tables:

exibble %>% gt()

Great Looking Tables: gt (v0.2)

Just like how the ggplot() function is the entry point to ggplot2 plots, the gt() function serves as the first function to call for making gt tables.

Formatting data in columns

The exibble dataset is blessed with an array of column types. This makes it a snap to experiment with gt ’s collection of fmt_*() functions, which format the input data values.

Let’s test as many formatter functions as possible. Here’s the plan:

  • have num display numbers with exactly 2 decimal places using fmt_number()
  • show nicely formatted dates in date using date_style 6 (the m_day_year style) with fmt_date()
  • format the 24-h time values in time to time_style 4 (the hm_p style) with fmt_time()
  • make the datetimes in datetime formatted as such with the fmt_datetime() function
  • transform the currency column with fmt_currency() to show us values in the euro currency ( currency = "EUR" )

Phew! Here’s the code and the corresponding gt table:

exibble %>%
  gt() %>%
  fmt_number(columns = vars(num), decimals = 2) %>%
  fmt_date(columns = vars(date), date_style = 6) %>%
  fmt_time(columns = vars(time), time_style = 4) %>%
  fmt_datetime(columns = vars(datetime), date_style = 6, time_style = 4) %>%
  fmt_currency(columns = vars(currency), currency = "EUR")

Great Looking Tables: gt (v0.2)

As can be seen, entire columns had formatting applied to them in very specific ways. There is some finer control available as well. We can style a subselection of rows in any given column and there are quite a few ways to specify the target rows (e.g., row indices, row names in the stub, conditional statement based on column data, etc.).

This only scratches the surface of what is possible in formatting the table body , there are more fmt_*() functions. If they don’t exactly suit your needs you can use the general fmt() function and provide your own transformation function.

A table with a header and a footer

We can add components to the table. Let’s include a header with a title and subtitle, and, a footer with a source note. These parts are added with the tab_header() and tab_source_note() functions.

exibble %>%
  gt() %>%
  tab_header(
    title = md("This is the `exibble` dataset in **gt**"),
    subtitle = "It is one of six datasets in the package"
  ) %>%
  tab_source_note(md("More information is available at `?exibble`."))

Great Looking Tables: gt (v0.2)

Adding new parts to the table is typically done by using a few tab_*() functions. Notice that we could style our text using Markdown with the included md() function.

Adding a stub and organizing rows into row groups

The exibble dataset has the row and group columns, which were purposefully included for experimentation with the table stub and with row groups . Rather than explaining those components at length, let’s revise the above code so that these columns are used to create those components:

exibble %>%
  gt(rowname_col = "row", groupname_col = "group") %>%
  tab_header(
    title = md("This is the `exibble` dataset in **gt**"),
    subtitle = md("We can use the `row` and `group` columns to structure the table")
  ) %>%
  tab_source_note(md("More information is available at `?exibble`."))

Great Looking Tables: gt (v0.2)

This change effectively gives us row labels in a separate area to the left (the stub ), and, row group labels above each grouping of rows. This is great for data that naturally falls into groupings. And worry not, if the initial order isn’t what you expected or wanted, the row_group_order() function can be used to reorder the groupings.

Using spanner column labels

Just as with the stub , we can create groupings of columns with spanner column labels that encompass one or more columns. The tab_spanner() function makes this possible. By providing a label and a selection of columns the new label is placed above those columns and the associated horizontal rule will span across. Should the columns not be adjacent to each other, tab_spanner() will automatically gather them together.

exibble %>%
  gt(rowname_col = "row", groupname_col = "group") %>%
  tab_spanner(label = "Dates and Times", columns = matches("date|time")) %>%
  tab_header(
    title = md("This is the `exibble` dataset in **gt**"),
    subtitle = md("We can use the `tab_spanner()` function to organize and label columns")
  ) %>%
  tab_source_note(md("More information is available at `?exibble`."))

Great Looking Tables: gt (v0.2)

More… so much more

It’s really not possible to explore much of what gt can do in a short blog post. You can do many more useful things like inserting footnotes, modifying text, borders, and fills, and, adding summary rows. Here’s an example of how the pizzaplace dataset can look with a little gt code (not shown here but available in this gist ):

Great Looking Tables: gt (v0.2)

Getting started with gt can be a risk-free experience with the gt Test Drive . Hit the button below to be transported to an RStudio Cloud project with examples galore:

Great Looking Tables: gt (v0.2)

To make it easy to experiment with making gt tables, we includedsix datasets in the package: countrypops , sza , gtcars , sp500 , pizzaplace (your favorite), and exibble . Strangely enough, each of these datasets is celebrated with a circular logo.

Great Looking Tables: gt (v0.2)

(Each of these datasets has a unique story in the world of gt so the deluxe graphics are warranted.)

While we’re only getting started on this package we feel things are really coming along. But sure to visit and engage with us at the gt issue tracker . We want to hear of any bugs, usage questions, or great ideas you might have to make this package better. Thanks!


以上所述就是小编给大家介绍的《Great Looking Tables: gt (v0.2)》,希望对大家有所帮助,如果大家有任何疑问请给我留言,小编会及时回复大家的。在此也非常感谢大家对 码农网 的支持!

查看所有标签

猜你喜欢:

本站部分资源来源于网络,本站转载出于传递更多信息之目的,版权归原作者或者来源机构所有,如转载稿涉及版权问题,请联系我们

程序算法与技巧精选

程序算法与技巧精选

郭继展 / 2008-5 / 36.00元

《信息科学与技术丛书•程序算法与技巧精选》分17章,139个例题。书中介绍的算法和技巧涉及到随机数函数理论,基础数论,新意幻方,提高程序运行速度和精度,特定数据排序,穷举、递推、递归和迭代等诸多方面。这些算法和技巧大多是作者历年从事教学、软件开发、学术研究和学习的成果总结。 《信息科学与技术丛书•程序算法与技巧精选》内容不涉及计算机专业课程的诸多概念、理论,读者只需要学过C语言,有算法、结构......一起来看看 《程序算法与技巧精选》 这本书的介绍吧!

JSON 在线解析
JSON 在线解析

在线 JSON 格式化工具

URL 编码/解码
URL 编码/解码

URL 编码/解码