he estimated baselines with two different regularisation constant values for the

a shown in Figure 5.5 using the baseline package.

e 5.9 shows the use of this package for the spectra data shown in

5 with two different regularisation constant values. It can be seen

n the regularisation constant changes, the estimated baselines

ferent shapes. As shown in the upper panel of Figure 5.9, the

d baseline was reasonable when the regularisation constant was

wever, the estimated baseline as shown in the lower panel Figure

worse when the regularisation constant was 15, where the

d baseline became a straight line. It was definitely useless for

ng signals from such a baseline because both the false peak

y rate and the missing true peak discovery rate will be

ably high.

e 5.10 shows the peak spectra extracted based on the estimated

as shown in Figure 5.9 using the baseline package, where two

regularisation constant values were employed. It can be seen that

regularisation constant was one, the extracted peak spectrum

ated a better pattern as shown in the upper panel of Figure 5.10.

k spectrum is certainly better for signal discovery. When the

ation constant was 15, the extracted peak spectrum failed to show

ttern as shown in the lower panel of Figure 5.10. Therefore signal

y from such a peak spectrum won’t be very successful.