---
title: "Formattable data frame"
author: "Kun Ren"
date: "`r Sys.Date()`"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{Formattable data frame}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r, echo=FALSE, include=FALSE}
set.seed(123)
```

Formattable data frames are data frames to be rendered as HTML table with formatter functions applied, which resembles conditional formatting in Microsoft Excel. 

## Simple examples

Suppose we have the following data frame:

```{r}
scores <- data.frame(id = 1:5,
  prev_score = c(10, 8, 6, 8, 8),
  cur_score = c(8, 9, 7, 8, 9),
  change = c(-2, 1, 1, 0, 1))
```

In the console, it is printed as plain texts:

```{r}
scores
```

Using `knitr::kable()` or `formattable()`, the data frame can be rendered as HTML table which looks more friendly.

```{r}
library(formattable)
formattable(scores)
```

In fact, `formattable()` calls `knitr::kable()` internally to translate data frame to HTML code. In addition, `formattable()` supports formatter functions to customize the transformation between values in the data frame to HTML code to generate.

```{r}
plain_formatter <- formatter("span")
plain_formatter(c(1, 2, 3))
```

```{r}
width_formatter <- formatter("span",
  style = x ~ style(width = suffix(x, "px")))
width_formatter(c(10, 11, 12))
```

The values of `change` can be positive, negative or zero. We can make positives green, negatives red, and zeros black by creating a formatter function that performs conditional transformation from value to HTML code.

```{r}
sign_formatter <- formatter("span", 
  style = x ~ style(color = ifelse(x > 0, "green", 
    ifelse(x < 0, "red", "black"))))
sign_formatter(c(-1, 0, 1))
```

Note that we don't have to write HTML but use helper functions like `style()` and `ifelse()` to make it easier to specify conditions. Then we call `formattable()` on the data frame with a list of formatter functions so as to apply conditional formatting.

```{r}
formattable(scores, list(change = sign_formatter))
```

We can also create another formatter function that makes above-average values bold while leaving others unchanged.

```{r}
above_avg_bold <- formatter("span", 
  style = x ~ style("font-weight" = ifelse(x > mean(x), "bold", NA)))
formattable(scores, list(
  prev_score = above_avg_bold,
  cur_score = above_avg_bold,
  change = sign_formatter))
```

## Cross formatting

Sometimes, we need to format one column based on the values of another column. This can be easily done with one-sided formula in `formatter()`. When using `formatter("span", style = ~ expr)`, `expr` is evaluated in the data frame so that all columns are available for use.

```{r}
formattable(scores, list(
  cur_score = formatter("span", 
    style = ~ style(color = ifelse(change >= 0, "green", "red")))))
```

## Hiding columns

To hide columns, use `FALSE` formatter.

```{r}
formattable(scores, list(prev_score = FALSE))
```

## Using built-in formatters

To making formatting easier, formattable package provides a group of built-in formatter functions. Suppose we have the following data on a number of products. Some columns are already formattable vectors.

```{r}
products <- data.frame(id = 1:5, 
  price = c(10, 15, 12, 8, 9),
  rating = c(5, 4, 4, 3, 4),
  market_share = percent(c(0.1, 0.12, 0.05, 0.03, 0.14)),
  revenue = accounting(c(55000, 36400, 12000, -25000, 98100)),
  profit = accounting(c(25300, 11500, -8200, -46000, 65000)))
products
```

Without any formatter functions applied, the formattable data frame is directly rendered as an HTML table.

```{r}
formattable(products)
```

We can supply a list of formatter functions to make it look more colorful. For example, we apply `sign_formatter` to `profit` column so that values of different signs are displayed in different colors.

```{r}
formattable(products, list(profit = sign_formatter))
```

Using built-in functions like `color_tile()` and `color_bar()` makes it easier to compare the magnitute of values of specified columns.

```{r}
formattable(products, list(
  price = color_tile("transparent", "lightpink"),
  rating = color_bar("lightgreen"),
  market_share = color_bar("lightblue"),
  revenue = sign_formatter,
  profit = sign_formatter))
```

## Area formatting

Sometimes, it is useful to apply a formatter function to an area so that all cells in the area share one benchmark. Area formatting is supported through the syntax of `area(row, col) ~ formatter` in the formatter list.

The following example renders the three columns altogether so that they share the same benchmark, not independently.

```{r}
set.seed(123)
df <- data.frame(id = 1:10, 
  a = rnorm(10), b = rnorm(10), c = rnorm(10))
formattable(df, list(area(col = a:c) ~ color_tile("transparent", "pink")))
```

If a one-sided formula is supplied, the function will be applied to all cells.

```{r}
formattable(df[, -1], list(~ percent))
```

# Dynamically generating formatters

Since `formattable()` accepts a list of formatter functions, the list can be dynamically generated. For example, the following code applies row-wise formatting, that is, each row is colored independently.

```{r}
df <- cbind(data.frame(id = 1:10), 
  do.call(cbind, lapply(1:8, function(x) rnorm(10))))
formattable(df, lapply(1:nrow(df), function(row) {
  area(row, col = -1) ~ color_tile("lightpink", "lightblue")
}))
```

# Converting to `DT::datatables`

`as.datatable()` is designed to convert a formattable data frame to `DT::datatables`.

```{r, screenshot.force = FALSE}
as.datatable(formattable(products))
```

Some formatters can be preserved well after the conversion.

```{r, screenshot.force = FALSE}
as.datatable(formattable(products, list(
  price = color_tile("transparent", "lightpink"),
  revenue = sign_formatter,
  profit = sign_formatter)))
```