Use R to Write LaTeX Code in R Markdown Documents
In latexSymb: R Functions for Readable LaTeX Mathematical Expressions

Technical reports written in R Markdown frequently incorporate mathematical symbols. R Markdown supports this requirement through the inclusion of raw LaTeX code, which is rendered by an appropriate LaTeX engine based on the desired output format. While this functionality is valuable, it presents certain challenges.

Writing LaTeX code can be cumbersome and often results in poor readability. While newcommand can mitigate these issues, it is underutilized in practice. This package leverages the fact that R Markdown users are already working within the R environment, enabling them to generate LaTeX code programmatically using familiar R syntax.

Example

The following example demonstrates this approach. Consider the following equation, encountered in Riemannian geometry:

$$ \frac{d}{dt} \left \langle V,W \right \rangle = \left \langle \frac{DV}{dt}, W \right \rangle + \left \langle V, \frac{DW}{dt} \right \rangle $$

The traditional LaTeX implementation would be:

\frac{d}{dt} \left \langle V,W \right \rangle = \left \langle \frac{DV}{dt}, W \right \rangle + \left \langle V, \frac{DW}{dt} \right \rangle

Using latexSymb, the same equation can be constructed programmatically. Begin by creating objects of class latex_symb to represent the vector fields:

library(latexSymb)
data(common)
attach(common)
vf1 <- lsymb("V")
vf2 <- lsymb("W")

Next, define R functions to represent the mathematical operations: the inner product, the covariant derivative, and the ordinary derivative:

inner <- function(x, y) ang(lsymb(x, ",", y))
cov.der <- function(x) lsymb("D", x) / "dt"
ddt <- function(x) lsymb("d", x) / "dt"

Finally, combine these elements within an equation environment:

lenv(
  "equation",
  lsymb(
    ddt(inner(vf1, vf2)),
    eq,
    inner(cov.der(vf1), vf2) + inner(vf1, cov.der(vf2))
  )
) |> cat()

Note: Use cat rather than print to prevent the display of escaped backslashes.

Implementation Details

This approach utilizes several key features of the latexSymb package.

The definition of cov.der demonstrates operator overloading in latexSymb. The / operator is redefined such that when either argument is a latex_symb object, it returns a new latex_symb object using the appropriate \frac command. Similar overloading is implemented for +, -, *, and ^ operators. The under function provides subscript functionality.

In the numerator of cov.der, lsymb("D", x) concatenates the string "D" with the object x, returning a new latex_symb object.

The ang function automatically handles delimiter sizing for angle brackets, eliminating the need for explicit \left and \right commands. Similar functionality is provided by br (braces), sqbr (square brackets), and pths (parentheses).

The lenv function simplifies environment creation by automatically generating \begin and \end statements. The second argument accepts a list or vector where each element represents a separate line. For example, the equation could alternatively be formatted as:

lenv(
  "align*",
  c(
    lsymb(ddt(inner(vf1, vf2)), "&=\\\\"),
    lsymb("&=", inner(cov.der(vf1), vf2) + inner(vf1, cov.der(vf2)))
  )
) |> cat()

The package also provides the il function, which wraps expressions in dollar signs for inline mathematical content.

Discussion

While latexSymb may not substantially reduce code volume compared to raw LaTeX, it enhances readability and reduces error likelihood. The package enables users to express mathematical concepts semantically rather than syntactically.

The package is not restricted to R Markdown and may be particularly well-suited for .Rtex documents. It is compatible with any format supported by knitr.

For complex equations, R's pipe operator can further enhance code organization and readability.