Why use knitR?

This is an R Markdown document, which present the basis elements to work with the knitr package in the goal to create report. Here, we presente a collection of tricks to work with knitR in R studio.

Why use knitR?

Generates report in HMTL, PDF, DOC
allows the reproductibility of the result
more easier that laTeX to produce documentation
shortcus integrated in R studio
combines all elements: documentation, code, result (figure, table, summaries)
is powerful (loops, interactivity, variability of formats, …)
uses Rmarkdown language

Rmarkdown language.

as HTML but without HTML tag
R code inclusion

Rmarkdown commands.

alt text

Create a new Rmarkdown document.

alt text

Include a R code in Rmarkdown document: chunk

R code is executed in a chunk
variables are global
chunk could have many options (or not)

Chunk options.

Many options are available to control the R code execution

alt text

Three ways to include a table.

with ktable.

# create a data.frame structure.
library(knitr)
mydata = data.frame(value1= 1:4, value2= 6:9)
rownames(mydata) <- paste("gene", 1:4)

kable(mydata, include.rownames=TRUE)

	value1	value2
gene 1	1	6
gene 2	2	7
gene 3	3	8
gene 4	4	9

with googlevis.

library(googleVis)

## 
## Welcome to googleVis version 0.5.8
## 
## Please read the Google API Terms of Use
## before you start using the package:
## https://developers.google.com/terms/
## 
## Note, the plot method of googleVis will by default use
## the standard browser to display its output.
## 
## See the googleVis package vignettes for more details,
## or visit http://github.com/mages/googleVis.
## 
## To suppress this message use:
## suppressPackageStartupMessages(library(googleVis))

mydata = data.frame(value1= 1:100, value2= 6:105)
rownames(mydata) <- paste("gene", 1:100)

table = gvisTable(mydata,options=list(page='enable', height='automatic', width='automatic'))
table

<!DOCTYPE html PUBLIC “-//W3C//DTD XHTML 1.0 Strict//EN” “http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd”> TableID2c2966e993aa

<!– Table generated in R 3.1.1 by googleVis 0.5.8 package –>

Data: mydata • Chart ID: TableID2c2966e993aa • googleVis-0.5.8

R version 3.1.1 (2014-07-10) • Google Terms of Use • Documentation and Data Policy

With DT package.

library(DT)

table = datatable(mydata, class = 'cell-border stripe')
table

<!–html_preserve–>

<!–/html_preserve–>

Include a plot in the report.

2 strategies to do that:

from a R code directly

Here, we precise no option for the R chunk {r}

mydata =rnorm(1000)
plot(mydata, cex= 0.4)

plot of chunk unnamed-chunk-4

In this case, we precise several options in the features of the plot : {r, fig.height= 14, fig.width= 14, fig.align='center'}

mydata =rnorm(1000)
plot(mydata, cex= 0.4, col= "red")

plot of chunk unnamed-chunk-5

from a file

R code to generate and save the plot

png("figure/plot2insert.png")
plot(mydata, cex= 0.4, col= "blue")
dev.off()

## png 
##   2

Include the plot with the markdown syntax.
alt text

Strategies with big Data or big Computation.

In the case, where you have a large dataset to analyze, it's important to review the way to process the data: think to process your computation on the side and save the result as a R object, which will be loaded in R chunks. Ideally, you create a R chunk with eval=F option to perform the computation: this R chunk is a component of your pipeline and is saved and available in the report ( its goal is just to generate the R objects to be visualized in the report without to be evaluated). It's a good habit to execute this part of code in Unix terminal and not in R studio server.For exemple, a function like mclapply() has a weird behaviour and batchjob usage doesn't work.

Conclusions.

very easy to use
Many functions are available in R studio interface
One document for script, documentation, result, reproducible
allows the interactive documentation with shiny package
allows an interactive data representation

The next time, to go further.

YAML header
CSS style
Rpres
loops and template
reference and bibliography
data representation with googlevis, qtlcharts and ggbio

SessionInfo

sessionInfo()

## R version 3.1.1 (2014-07-10)
## Platform: x86_64-redhat-linux-gnu (64-bit)
## 
## locale:
##  [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
##  [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8    
##  [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
##  [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
##  [9] LC_ADDRESS=C               LC_TELEPHONE=C            
## [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       
## 
## attached base packages:
## [1] stats     graphics  grDevices utils     datasets  methods   base     
## 
## other attached packages:
## [1] DT_0.1          googleVis_0.5.8 knitr_1.10.5   
## 
## loaded via a namespace (and not attached):
##  [1] digest_0.6.8     evaluate_0.7     formatR_1.2      highr_0.5       
##  [5] htmltools_0.2.6  htmlwidgets_0.4  jsonlite_0.9.16  magrittr_1.5    
##  [9] RJSONIO_1.3-0    rstudioapi_0.3.1 stringi_0.4-1    stringr_1.0.0   
## [13] tools_3.1.1      yaml_2.1.13