x[which(x[,1]>0),1])

,nclass=50,xlab='log(count)',

main='SRR1039508',prob=TRUE)

density(z))

he histogram of the logarithm-transformed non-zero sequencing counts of the

R1039508 with a smooth density curve added for the airway data.

mportant issue of the histogram approach is the proper design of

umber. When employing too few bins, a density function may

more flattened density curve. Figure 2.6(a) shows such a case,

ly four bins were employed. When employing too many bins, a

unction is very non-smooth, which may not be able to reveal a

butional pattern. Figure 2.6(c) shows such a case, where 200 bins

ployed. Only a proper bin number can help accurate pattern

y. Figure 2.6(b) shows a case, where 20 bins were employed and

ty pattern can be easily visualised for the data.

(a) (b) (c)

he impact of the bin number on density quality for a normally density. (a) The

with too few bins. (b) The histogram with a proper number of bins. (c) The

with too many bins.