RMarkdown & Knitr

Rollie Parrish
Dec 5, 2014

Overview

  • R Markdown used to generate dynamic documents from R
  • Depends on:
    • Rmarkdown & knitr packages
    • Pandoc & RStudio

R Markdown

Basic Markdown

  • narrative content written in plain text
  • basic formatting using a simple syntax
  • Markdown is commonly used in web content (blogging, Wikipedia content, etc.)

R code chunks

  • incorporates the results of R code

Rendered into final output

  • html / pdf / Word / Presentations (like this one)

Inline Example

### Results
2 + 2 = `r 2 + 2`

Results

2 + 2 = 4

Knitr

Processes the R Code chunks along with R Markdown to produce final output

“Chuck options” control how the R code is processed.

Example: Graphics

{r plot1, fig.width=10, fig.align='center'}

plot of chunk plot1

Example: Graphics

{r plot2, fig.width=5, fig.align='right'}

plot of chunk plot2

A More Elaborate Example

Question Is the percentage of smokers significantly different between groups of patients?

Data The count of patients and the count of smokers.

conf.level <- 0.99
smokers  <- c( 80, 84)#, 129, 90)
patients <- c( 86, 93)#, 136, 120)

A More Elaborate Example

results <- prop.test(smokers, patients)
results


    2-sample test for equality of proportions with continuity
    correction

data:  smokers out of patients
X-squared = 0.1456, df = 1, p-value = 0.7028
alternative hypothesis: two.sided
95 percent confidence interval:
 -0.06486486  0.11887836
sample estimates:
   prop 1    prop 2 
0.9302326 0.9032258

A More Elaborate Example

results are assigned to variables so we can insert the elements inline

estimates <- paste(round(results$estimate * 100,1),"%",
                   collapse=" vs. ", sep="")
p_value <- ifelse(results$p.value < .001,
                  0.001, round(results$p.value,3))
p_eq_lt <- ifelse(p_value > .001,
                  "=", "<")
p_sig <- ifelse(p_value > 1-conf.level,
                "not", "")

Results

p_value p_eq_lt p_sig estimates
0.703 = not 93% vs. 90.3%

A More Elaborate Example

This example evaluates the smoking status from `r length(patients)` groups of patients.

Chi-square analysis indicates the proportions of smokers are `r p_sig` significantly different between groups
(`r estimates`, $p$-value `r paste(p_eq_lt, p_value)`).

This example includes the smoking status from 2 groups of patients. Chi-square analysis indicates the proportions of smokers are not significantly different between groups (93% vs. 90.3%, \( p \)-value = 0.703).

Additional Resources