Readme.Rmd

---
title: "Readme"
author: "Anton Könneke"
date: "2023-06-01"
output: github_document
---

This repository stores some helper functions that I use.

# Save ggplot as tikz

This is handy for exporting ggplots to LaTeX with the `ggsave` framework and `last_plot`. 
The function builds on the approach outlined by [Ilyass Tabiai](https://github.com/ilyasst) in this [blog post](http://iltabiai.github.io/tips/latex/2015/09/15/latex-tikzdevice-r.html). The main part is done by the 
package [tikzDevice](https://github.com/daqana/tikzDevice) by Charlie Sharpsteen et al.

```{r}
library(ggplot2)
library(tikzDevice)
source("ggtiksave.R")

ggplot(mpg)+
  geom_point(aes(displ, cty, color = drv))+
  theme_minimal()+
  xlab("Look, greek letters $\\alpha \\beta \\Gamma \\Phi$")+
  ylab("And Math! $\\frac{20}{19} \\prod_{k=1}^n$ ")+
  ggtitle("Wait, is this \\LaTeX?")

ggtiksave("myplot.tex")
```
And this is the result if `myplot.tex` is `input` into a tex doc (with tikz loaded!).

![Latin Modern](latex_screenshot.png)

The neat thing is that now, whenever the font of the \Latex file is changed, so 
is the font in the plot:

## Palatino Font
![Palatino font](font1.png)

## Linux Libertine Font
![Libertine font](font2.png)

# Encode Excel columns
Quite frequently when working with human generated excel files, 
spaces are accidentially stored in numeric cells. In addition to that, it is 
common to use `,`instead of `.` as decimal markers in germany. As neither
`readxl::read_xlsx` nor `openxlsx::read.xlsx` seem able to account for this, this little function 
does the correction afterwards. By wrapping it in `mutate(across())` call, the job can be easily done.

```{r message=FALSE}
source("fix_excel_numcols.R")
require(dplyr)
df <- tibble(
  A = c("1,3", "1,5 ", "2"),
  B = c("1,1", "5,5 ", " 2,4")
)

df %>% 
  mutate(across(everything(), ~fix_excel_numcols(.x)))
```