Skip to main content
. 2019 Sep 20;9:13644. doi: 10.1038/s41598-019-50178-0

Table 1.

Choose the best number of profiles for day 1.

Number of profiles LL AIC BIC aBIC Entropy AICC P* Number of patients in each profile (%)
1 2 3 4 5 6 7
2 −96714.16 193562.3 193911.3 193698.5 0.902 193569.4 0.2230 241 (18) 1111 (82)
3 −95836.17 191852.3 192321.2 192035.3 0.945 191865.3 0.0140 175 (13) 208 (15) 969 (72)
4 −95366.65 190959.3 191548.0 191189.0 0.954 190980.1 0.4395 25 (2) 1100 (81) 220 (16) 7 (1)
5 −94777.82 189827.6 190536.1 190104.1 0.959 189858.3 0.7600 172 (13) 76 (6) 905 (67) 190 (14) 8 (1)
6 −94433.30 189184.6 190012.9 189507.8 0.960 189227.3 0.0125 1 (0) 209 (15) 1025 (76) 7 (1) 85 (6) 25 (2)
7 −93791.13 187946.3 188894.4 188316.2 0.968 188003.3 0.1260 164 (12) 22 (2) 933 (69) 154 (11) 71 (5) 7 (1) 1 (0)

*P value was reported comparing k-profile model to (k-1)-profile model based on the VUONG-LO-MENDELL-RUBIN likelihood ratio test.

Abbreviations: AIC: Akaike Information Criterion; AICC: Akaike Information Criterion corrected; BIC: Bayesian information criteria; aBIC: adjusted Bayesian information criteria.