Setting a scale in ggplot2

scale_y_continuous(limits = c(10, 200))

# Month: December 2010

## Moving average in R Median: runmed() M …

Moving average in R

Median: runmed()

Mean: filter()

ex)

runmed(data, 3)

filter(data, rep(1/3, 3))

## Save ggplot in pdf Use dev.copy2pdf(fil …

Save ggplot in pdf

Use dev.copy2pdf(file=”plot.pdf”)

Example)

plotA <- ggplot() + geom_line(data=nucleosome.mono.colsum, aes(seq, y=A, colour="A") )
plotT <- ggplot() + geom_line(data=nucleosome.mono.colsum, aes(seq, y=T, colour="T") )
arranged.plot <- grid.arrange(plotA, plotT, nrow=2)
dev.copy2pdf(file="arranged.plot")
dev.off()

## # Running R in batch mode R CMD BATCH R …

# Running R in batch mode

R CMD BATCH Rscript.R

## Object in R Factor can be used for sequ …

Object in R

Factor can be used for sequence data.

From sequence characters

seq <- c("A", "C", "C", "T", "G")
seq.factor <- factor(seq, levels=c("A", "G", "T", "C")
From numbers to sequences
> seq.num <- c(1,2,3,1,1,3,2,1,4,1,3,2,1)
> seq.num.factor <- factor(seq.num, levels=c(1,2,3,4), labels=c("A","G","T","C") )

## There seems to be a bug in a R package n …

There seems to be a bug in a R package named pgam:periodogram().

The function intensity happened to have two cos. One of them should be sin.

Here is the correct version.

function (w, x)

{

n <- length(x)
t <- seq(1:n)
sp <- ((sum(x * cos(w * t)))^2 + (sum(x * sin(w * t)))^2)/n
return(sp)
}

## R vector does not have dimension. It onl …

R vector does not have dimension. It only has length.

On the other hand, matrix has dimension.

## R commands to explore time series data …

R commands to explore time series data

ts(data) # create time series objects

time(ts.data) # returns the time

diff(ts.data, lag=1) # difference at the specified lag

lag.plot(ts.data, 9, do.lines=FALSE) # plot with specified lags

plot(stl(ts.data, “per”)) # decomposition of time series data

spec.pgram(ts.data, spans=c(3,3), log=”no”) # non-parametric spectral analysis

spec.ar(ts.data, spans=c(3,3), log=”no”) # parametric estimation

http://www.stat.pitt.edu/stoffer/tsa2/R_time_series_quick_fix.htm

## R commands examining data and its struct …

R commands examining data and its structures

ls() or objects() # list variables

structure()

str() # summary of data

levels()

dim() # dimension of data

class() # show the class of the variable

head() or tail() # showing parts of data

http://www.statmethods.net/input/contents.html

## R normality tests shapiro.test() his …

R normality tests

shapiro.test()

hist(data, prob=TRUE)

lines(density(data))

qqnorm(data)

qqline(data)