y

Gaussian distribution with a zero mean and a unit standard

. The population statistics (the mean and the standard deviation)

Figure 2.1 are all slightly different from the expectation. One of

rtant objectives of density estimation is thus to estimate a density

for a data set as accurately and as precisely as possible. A density

can be estimated using several approaches. The following

introduce several commonly used ones as well as their

ntations in the R environment.

he estimated densities from a number of collected data sets for an expected

istribution with a zero mean and a unit standard deviation. The solid line stands

density. The dotted lines stand for the densities estimated based on the samples

m the true distribution with smaller sizes. In the legends, ‘mu’ stands for the

means and ‘sigma’ stands for the population standard deviations.

e histogram approach

ogram approach is the simplest approach for density estimation

1985]. Its basic principle is to count data points which fall in

re-designed bins and use these counts to profile a density function

set.

e 2.2 shows how a histogram is constructed for the gene isftu1 of

cisella Tularensis species in a gene essentiality pattern analysis

Yang, et al., 2017]. The gene attracted 7,753 transposon insertions.

sertions were distributed across 180 sites within the gene. The

se pairs were between 3,204 and 3,988 in the genome.

poson insertion sites were between 3,207 and 3,980 base pairs.