Table 5 Best fitting distributions for each continuous variable, without and with Kolmogorov–Smirnov outlier removal, considering 5 categories per variable
From: DI2: prior-free and multi-item discretization of biological data and its applications
Variables | Without opt. | \({\tilde{\chi }}^2\) statistic | p-value >0.05 (\({\tilde{\chi }}^2\)) | D statistic | With opt. | \({\tilde{\chi }}^2\) statistic | p-value >0.05 (\({\tilde{\chi }}^2\)) | D statistic |
|---|---|---|---|---|---|---|---|---|
mcg | foldcauchy | 3.72 | True | 0.08 | exponnorm | 3.18 | True | 0.02 |
gvh | genlogistic | 3.57 | True | 0.03 | genlogistic | 2.02 | True | 0.02 |
alm | genlogistic | 17.00 | False | 0.05 | genlogistic | 12.08 | False | 0.03 |
mit | exponnorm | 19.23 | False | 0.05 | exponnorm | 6.11 | True | 0.03 |
pox | chi2 | \(4.4 \times 10^{-14}\) | True | 0.99 | gengamma | \(4.2 \times 10^{-14}\) | True | 0.99 |
vac | laplace | 20.99 | False | 0.08 | pearson3 | 14.18 | False | 1.00 |
nuc | exponnorm | 1116.63 | False | 0.26 | mielke | 795.28 | False | 0.26 |
all variables | genhalflogistic | 45.69 | False | 0.25 | genhalflogistic | 10.25 | False | 0.21 |