Table 1. Different ranking of the sample table of sensitivity indices.
A. | Sensitivity values (raw) | B. | Simple ranks (SR) | ||||||
M1 | M2 | M3 | M4 | M1 | M2 | M3 | M4 | ||
K1 | 100 | 45 | 99 | 10 | K1 | 4 | 2 | 3 | 1 |
K2 | 100 | 55 | 20 | 30 | K2 | 4 | 3 | 1 | 2 |
K3 | 98 | 34 | 45 | 100 | K3 | 3 | 1 | 2 | 4 |
K4 | 85 | 85 | 79 | 82 | K4 | 3 | 4 | 1 | 2 |
K5 | 50 | 30 | 20 | 47 | K5 | 4 | 2 | 1 | 3 |
K6 | 0 | 0 | 0 | 0 | K6 | 1 | 2 | 3 | 4 |
– parameters K4 and K6 seem to have similar sensitivity, however K4 is highly sensitive (according to raw values from A), and K6 is not sensitive at all; D. improved average ranks (IAR) of A – here the K4 is shown to be sensitive, and K6 is insensitive. Model states are located in columns (M1 to M4), parameters are located in rows (K1 to K6). A. Sample raw values of SIs; B. simple ranks (SR) of A - direct indication of sensitive parameters (rows) is not feasible; C. average ranks (AR) of A