R_commands

OVERVIEW OF R COMMANDS

UPDATING R
zypper update R-patched (while superuser)

GENERAL
?Rcommand	
or
help("Rcommand")	Returns help page of the Rcommand
example(Rcommand)	Shows you examples of Rcommands
quit() or q()		Exits R command line
Vector 	  		An R object that holds a 1 dimensional list of values or strings
Matrix			Table of values with rows and columns
List			An R object that is like a scaffold. 
			Can hold multile vectors, matrices, dataframes at the same time in the same list
class(object)		Will return to you the what kind of object type your object is
unclass(object)		Will break down to you how the object is build up
ls()			At any time, will list all of your current variables
install.packages()	Install a package
library()		Load a package
require("package")	Check if package is installed
R.version		List that displays the current version of R


DEFAULT VECTORS
letters			List of the alphabet
letters[2:5]		Will list b,c,d,e

NAMES
names(			Associates 'names' with vector elements
	vector,
) <- c('name1','name2','name3','etc')
names(<list>)		Returns the components of a list

EXPORTING A PLOT
png(
	filename="x.png",	"your/file/location/name.png"	Set path and format to save file to 
	width=480,		Set width of file 
	height=480,		Set height of file
	units="px"		Set the unit used by width and height
)
	
pdf(
	file="path/to/file.pdf",
	paper="a4"
)

Other options:
svg()
Then make the plot within R
After you're done, type
dev.off()

EXPORTING A VECTOR
write(
	x,				The vector
	file="",			Specifying file name
)

EXPORTING A TABLE
write.table(
	x,
	sep="\t",
	file="foo.table",
	quote=F,			If TRUE, any character will be surrounded by double quotes
	col.names=F,
)

IMPORTING A TABLE			Import a table into a data.frame
df<-read.table(
	file,				File that contains the table you want to import
	sep="\t",			State the separator that separates the columns in the file
	header=TRUE,			State if the table contains a header or not
	row.names=c()			Provide vector with row    names if you want to
	col.names=c()			Provide vector with column names if you want to
	row.names=1			If your table also contains row headers
	skip=<n>			Skips the first <n> lines before reading the data
)

DATAFRAMES
dataframe<-read.table("path/to/table")	 Generates dataframe from file
dataframe[rows,columns]
For example
dataframe[1,]				 Returns first row of dataframe
dataframe[,1]				 Returns first column of dataframe
dataframe$header			 Returns column with the header 'header'
row.names=1				 If your table also contains row headers,
					 Tells R that the first columns should be used as row names
dataframe[,c(1,3,5)]			 Returns only columns 1, 3, and 5 of dataframe
dataframe[ ,c(dog,cat,cow) ]		 Returns only colums with the headers 'dog', 'cat' and 'cow')
dataframe[dataframe$col>=1,]   		 Selects all rows of the dataframe for which the value in the column 'col'
					 is greater than or equal to 1.
df[df$col1>=1 & df$col2<=4, ]		 Selects all rows for which both conditions return TRUE
colnames(df)<-c("col1", "col2")		 Renames standard column names (V1 -> col1, V2 -> col2)
data.frame(				 Create a data.frame from scratch.
	col1=c("gras","sky","blood"),	 col1 contains values "gras", "sky" and "blood". you can rename col1 to whatever you want
	col2=c("green","blue","red"),	 col2 contains values "green", "blue", and "red". you can rename col2 to whatever you want
)

SETDIFF
Returns a substracted vector, where one vector is substracted from the other	      		    		 
setdiff(
	vector1,			The bigger vector
	vector2,			The smaller vector, which is a subset of the bigger vector
)
				 
HISTOGRAM
hist(
	dataframe[,1],			List of numbers for the histogram 
	xlim=range(start,end)		X-axis range
	breaks=number			Amount of bars in the histogram
	main="title"			Title of the plot
	xlab="title"			Title of x-axis			
	col="darkgreen"			Color of the bars
	add=T				If you want to overlay over an existing histogram
)

DENSITY PLOT (similar to histogram, but no bars, just a line)
plot(density(x))

SCATTER PLOT
plot(
	x=dataframe[,1]			List of x-values
	y=dataframe[,2]			List of y-values
	xlim=c(0,2000)			Set x-axis limits (when using log scale, dont use 0 as start value)
	ylim=c(0,5000)			Set y-axis limits
	xlab="name"			Name your x-axis
	ylab="name"			Name your y-axis
	log="y"				Transform y-axis into logarithmic scale
	log="xy"			Transforms both axes into logarithmic scale
	las=0,1,2,3			Horizontal y-axis numbers* 
)

You can also specify x and y values before you make the plot:
    	x <- list$x    	     	    	Pull x values into 'x'
	y <- list$y			Pull y values into 'y'
	plot(x,y)			Will plot x vs y  

PLOT AESTHETICS
par(
*las=0 (default)		Vertical y-axis, horizontal x-axis numbers
las=1  				Horizontal y, horizontal x
las=2				Horizontal y, vertical x
las=3				Vertical y, vertical x 
type=".."			Type of plot
type="p"			Points
type="l"			Line
type="b"			Both
type="n"			No plotting
pch="3"				The symbol of your datapoints. See R course documentation for different symbols
font.lab=2			Edits the font of the axis labels.
				1 = regular
				2 = bold
				3 = italics
				4 = bold italics
cex.axis=0.5			Scales the size of axis labeling text to provided factor on both axes
				For example, 0.5 means that text labels will be half the size compared to default
cex=0.3				Scales the sizes of the points in the plot to provided factor
				For example, 0.3 means that the points will be 0.3 times the size compared to default
mar=c(4,5,1,1)			A vector of size 4, providing the margins for the bottom left top and right
par()$mar			Current margin settings
xaxt="n"			Suppress the x-axis labeling
par(new=T)			State after plot and before plotting a new plot overlaying the previous one
lwd=1.5				Set line width
bty="n"				Disable box for plot (only draws 2 axes instead of 4)
mfrow=c(1,3)			Set up 3 spaces in a row where plots will be added when they are created
)
box()				Will create a box around your plot

CUSTOM AXIS
axis(
	side,			1 (bottom-horizontal-axis) or 2 (left-vertical-axis) or 4 (right-vertical-axis)
	at=vector,		vector containing locations of major ticks
	labels=vector,		vector containing which locations to label
	tcl=-0.25		for small ticks
)

ABLINE				Adds straight lines to plots
abline(
	v="x-value",		v for vertical
	h="y-value",		h for horizontal
	lty=2,			Dotted line
	col="darkgrey",
)

GRID
After you made the plot, you can add some stuff:
        grid() 	   	     	        Draws a light grid in the plot	  	        

MTEXT				Write text in one of the four margins of a plot
mtext(
	text="text",		Write "text" in the margin
	side="1|2|3|4",		1=bottom, 2=left, 3=top, 4=right
	line=<number>		Write "text" on the nth line, starting at 0, and increasing with 1 while going outwards
	col="col"		Color of the text
)
MATRIX
matrix(
	data="data vector"		List of values that you want to fill the matrix with
	nrow="number"			Desired number of rows 
	ncol="number"			Desired number of columns
	byrow=TRUE			If TRUE, matrix will be filled by rows
					If FALSE, matrix will be filled by columns
)
row.names(matrix)<-df$COG		Give row.names to the matrix

Example:
list <- read.table("file with list of values")
matrix(data=list[,1], nrow=16, ncol=24, byrow=TRUE)

You can also assign a dataholder to the matrix:
matrix <- matrix(data=list[,1], nrow=16, ncol=24, byrow=TRUE)

Rename the column and row headers:
colnames(matrixname) <- paste(1:24, sep="")
rownames(matrixname) <- paste(letters[1:16], sep="")

DIM()
dim(
	matrix				Returns the dimensions of a matrix (rows, columns)
)
DATA.MATRIX()
Converting dataframe to matrix
data.matrix(
	matrix,
) 

NCOL(), NROW()				Returns the number of columns or number of rows of a dataframe, matrix

IMAGE/CONTOUR/PERSP
Visualization of the matrix
image(					Creates a heatmap of the matrix (don't know how its different from heatmap)
	matrix,
)
contour(				Draws contour map of matrix. Like a relief map
	matrix,				
)
persp(					Draws a 3D image of the matrix
	matrix,
	expand=0.2,			Changes at which value you wish to center your view
)
	
HEATMAP
Plot a heatmap from a matrix

heatmap(
	matrix_name			The matrix that contains all the data for the heatmap
	Rowv=NULL			Determines whether or not you see a dendogram for the rows
					If NA, dendogram will be suppressed
					If NULL, dendogram will be displayed
	Colv=NULL			Same as Rowv, but for the columns
)
heatmap.2( ## Requires 'gplots' package ##
	matrix,
	Colv=T,
	Rowv=T,
	dendrogram="none",
	breaks=c(0,0.5,1.5,2.5,3.5),	Position of the color transitions in the color key
	col=colorpanel,			Selects color panel for the color key. Can be standards or customized
	trace="none",			Suppresses the trace line in the heatmap
	colsep=c(1:25),			Introduces seperator lines between the columns 1 till 25.
	sepwidth=c(0.001,0.001),	Sets the width of the seperator lines of (rows,columns)	
	sepcolor="black",		Sets the color of the seperator lines
	lmat=rbind(c(2,4), c(3,1)),	Sets the layout matrix of the plot
			   		1 = The heatmap
					2 = Row dendrogram
					3 = Column dendrogram
					4 = Color Key
	lhei=c(0.7,3),			Sets the height of the rows of the layout matrix (NOT of the heatmap rows)
	lwid=c(0.2,1),			Sets the width of the columns of the layout matrix
	keysize=0.01,			Sets the size of the color key
	density.info="none",		Suppresses the histogram inside the color key
	xlab="Taxa",			X-axis title
	ylab="RickCOGs", 		Y-axis title
	margins=c(5,5),			Margin sizes of column and row names, respectively
	cexCol=0.5,			Set font size of column labels to 0.5* the default
)

COLORPANEL()
colorpanel(
	4,				Number of colors you want in your colorpanel
	low="white",			Set your start color of your gradient
	high="darkblue",		Set your end color of your gradient
)
Standard R colorpanels:
	rainbow()
	heat.colors()
	terrain.colors()
	topo.colors()
	cm.colors()

RCOLORBREWER
library(RColorBrewer)
# Create a palette of 7 greens
# "Greens" and "Oranges" are preloaded palettes that come with RColorBrewer
my_palette<-brewer.pal(n=7, name="Greens")
# Create a palette of 7 greens, a grey, and 9 oranges
my_palette<-c(rev(brewer.pal(n=7, name="Greens")), "#D3D3D3", rev(brewer.pal(n=9, name="Oranges"))
# View all colorblind friendly palettes preloaded with RColorBrewer
display.brewer.all(colorblindFriendly=T)

ORDER
First sorts a numerical vector by ascending order. Then, in the sorted list, it will translate each element to the position in the original vector

EXAMPLE:
vector	5 52 7  13 22
sort    5 7  13 22 52
order   1 3  4  5  2		'5' is the 1st element in the orginal list, '7' the 3rd, '13' the 4th, etc.

Very handy when you want to sort a matrix or dataframe.

order(
	vector,				The vector that you want to sort
	decreasing=TRUE,		Whether you want to sort increasingly or decreasingly
)

Sort a dataframe example:
df[order(df[,6],decreasing=T),]		The largest value in df[,6] get index position 1 via order(). The  

FORMAT
Can be for example be used to force non-scientific notation in a plot.
format(c(20000,40000,60000,80000,100000),scientific=F)

APPLY
Applies a function to rows or columns of a matrix or dataframe or array
apply(
	df[,rick],		the dataframe or subset of dataframe
	1,			1 = rows, 2 = columns
	function(x)any(x>=1),	the function. Are any of the values in this particular row above 1?
				Returns TRUE for this row if condition is met
				Returns FALSE for this row if condition is not met
)

!apply()			Reverses FALSE and TRUE

BARPLOT
barplot(
	matrix,			Contains the data. Columns are the categories, rows the counts within those categories
	beside=TRUE,		TRUE -> Stacked barplot, FALSE -> Grouped barplot
	col=c("blue","red")	Colors of the bars,
	legend=TRUE,		Toggles legend
)

XTABS
Build a matrix from a two-dimensional data.frame.
For example, if data.frame is like: 

lake	     sag	   count
Spring	     B11	   105
Spring	     J10	   239
Spring	     L15	   163
Summer	     B11	   500
Summer	     J10	   1000
Summer	     L15	   750

You can build a matrix that looks like
    Spring    Summer
B11 105	      500
J10 239	      1000
L15 163	      750

xtabs(
	count~lake+sag, 
	df
)

MERGE
Merge two or more data.frames by common columns or row names
merge(
	df1,		Dataframe 1
	df2,		Dataframe 2
)

INTERSECT
Find the common elements between two vectors

vector1=c(1,2,3,4,5)
vector2=c(2,4,6,8,10)
intersect(vector1,vector2)	Returns "2", and "4"

TABLE
as.dataframe.table(<dataframe>)		Returns a new dataframe with a count for each unique element in 

SUBSET
Subset a dataframe or vector based on certain conditions
subset(
	x=<dataframe|vector>,		object to be subsetted
	subset,				logical expression indicating rows (dataframe) or elements (vector) to keep
)
Example:
markers_bin_187_subBin0<-subset(markers_bin_187, scaffold %in% bin_187_subBin0$scaffolds$scaffold)

SUM
You can use sum to count the number of times a given value occurs in a vector.
For example:
v<-c(0,0,1,1,1,1,2,2,2,2,3,3,4,4,5,5,5,5,5,5)
sum(v==0) returns 2
sum(v==1) returns 4
sum(v==2) returns 4
etc

PCA plot
library(ggplot2)

# convert to frequencies
df.freqs<-df.s[,2:21] / df.s$sum

# PCA plot
df.data contains the values that are used for PCA
df      contains more information, can be used to color points by group

autoplot(prcomp(df.data), data=df, colour="clade")
+ scale_color_manual(values=c("blue","red","orange","grey")

+ scale_color_manual is a general ggplot2 function you can use to manually provide your own colors

autoplot() is from the ggfortify package. Use the latest github version, it will show the explained variance.

GGPLOT2
gglot(data=dataframe, aes(x=column, y=column) + geom_point() + xlab() + ylab() + ggtitle()

# put multiple plots in a single page
require(gridExtra)
plot1 <- qplot(1)
plot2 <- qplot(1)
grid.arrange(plot1, plot2, ncol=2)

ggsave(filename="boxplot.pdf", plot=p, device=cairo_pdf, width=190, height=130, units="mm"
RESHAPE2
d:
    am18_f050 am18_untr am24_f050 am24_stat am24_untr nucl29_untr
1     47.29     71.51     73.92     10.32    115.45      346.89
2     83.06    146.21    185.82      8.74    231.87      110.61
3    104.56    179.87     82.42      5.96    127.96       50.89
4     41.08     64.16     53.86      1.94     91.73      207.82
5     63.15    144.22     35.58      5.65     45.68       57.38
6     37.20     49.67     32.20      5.94     40.48       43.00

melt(d, na.rm=T, variable.name="alignment", value.name="x2score")

alignment x2score
1 am18_f050   47.29
2 am18_f050   83.06
3 am18_f050  104.56
4 am18_f050   41.08
5 am18_f050   63.15
6 am18_f050   37.20
7 am18_untr   71.51
8 am18_untr   146.21
9 am18_untr   179.87
10 am18_untr  64.16
11 am18_untr  144.22
etc etc