Skip to main content

Advertisement

Springer Nature is making Coronavirus research free. View research | View latest news | Sign up for updates

Figure 6 | Source Code for Biology and Medicine

Figure 6

From: Using the value of Lin’s concordance correlation coefficient as a criterion for efficient estimation of areas of leaves of eelgrass from noisy digital images

Figure 6

Dependence of both the value of r for maximum ρ ^ r and the maximum value of ρ ^ r itself on sample size. For each value of the sample size index p = 1, 2, … 8 a fixed number s(p) of samples of size 100p each were uniformly drawn from the population of observed leaf areas. For each one of the samples in a set s(p), we iterated values of r through the interval 1 ≤ r ≤ 255, and for each one of these r values we obtained the concomitant concordance correlation coefficient values ρ ^ r . We repeated this procedure for all the samples in the set s(p) and averaged the r values at which ρ ^ r attained its maximum value, the obtained averages for the different values of the sample size index p are shown in a). The maximum values that ρ ^ r obtained in a sample were also averaged over the s(p) sets. These average values depending on sample size are correspondingly shown in b). The results of this study shows that neither the optimal interval of tolerance of similarity ST(r) or the reproducibility of observed leaf areas by means of their digital image surrogates depend on sample size, therefore the CCM can be considered as, a robust procedure.

Back to article page