[A-D 変換 PCI ボード]
1 枚 16 万円. なんでそんなに高いの? 甲山さん・小池さんといった 金づかいのアラいおじさんたちが 何か画策してるらしい …… いやはや. |
vi improved
こと vim
の小ワザひとつ
(vim
の :help 2html
に書いてあったもの).
vim -f +"syn on" +"run! syntax/2html.vim" +"wq" +"q" ファイル名
vim
は勝手に終了する).
生成例としては (いつものごとし,なんだけど)
こんなの
.
ただしこれは生成結果を Perl スクリプトにほうりこんで,
色を変えている.
#=== GOS00text ================================================== #--- Model E -------------------------------------------------- transfer_AA: 338 transfer_AC: 53 transfer_CA: 56 transfer_CC: 1082 transfer_total: 1529 prob_AA: 0.221 prob_AC: 0.035 prob_CA: 0.037 prob_CC: 0.708 conditional_prob_A: 0.142 conditional_prob_C: 0.047 plot = E model = E log_likelihood = -2.2619e+02 #--- Model M -------------------------------------------------- ... #--- Model E+M -------------------------------------------------- transfer_AA: 826 transfer_AC: 97 transfer_CA: 103 transfer_CC: 2437 transfer_total: 3463 prob_AA: 0.239 prob_AC: 0.028 prob_CA: 0.030 prob_CC: 0.704 conditional_prob_A: 0.111 conditional_prob_C: 0.038 plot = E model = E+M log_likelihood = -2.2962e+02 plot = M model = E+M log_likelihood = -2.1058e+02 #--- Model M+L -------------------------------------------------- ... #--- Model E+M+L -------------------------------------------------- transfer_AA: 2688 transfer_AC: 123 transfer_CA: 129 transfer_CC: 2653 transfer_total: 5593 prob_AA: 0.481 prob_AC: 0.022 prob_CA: 0.023 prob_CC: 0.474 conditional_prob_A: 0.046 conditional_prob_C: 0.044 plot = E model = E+M+L log_likelihood = -2.5387e+02 plot = M model = E+M+L log_likelihood = -2.1931e+02 plot = L model = E+M+L log_likelihood = -1.2359e+02
log_likelihood
を計算してる,
ということで.
likelihood_ratio_test
と
AIC
を計算させるようにようにする.
これで一件落着か.
今日はコードいじりやめ.
#=== GOS00text ================================================== # E,M,L v.s. E+M,L likelihood_ratio_test: df = 2 chi2 = 1.064e+01 prob = 0.0049 AIC model selection: E,M,L = 1.0520e+03 E+M,L = 1.0586e+03 model = E,M,L number_of_q[AC] = 6 log_likelihood = -5.1998e+02 model = E+M,L number_of_q[AC] = 4 log_likelihood = -5.2530e+02 model = M+L,E number_of_q[AC] = 4 log_likelihood = -5.6144e+02 model = L+E,M number_of_q[AC] = 4 log_likelihood = -5.8009e+02 model = E+M+L number_of_q[AC] = 2 log_likelihood = -5.9677e+02