Closed otknoy closed 8 years ago
20人分を一気に求めるコマンド
$ for i in `seq -w 20`; do node experiment/ec_exp.js ../ashidaResearch/experiment/subjects/experiment${i}.json > experiment/result/experiment${i}_result.csv
csv ファイルが生成される。
otknoy@debian:~/repos/leap_practice$ ls experiment/result/
experiment01_result.csv experiment04_result.csv experiment07_result.csv experiment10_result.csv experiment13_result.csv experiment16_result.csv experiment19_result.csv
experiment02_result.csv experiment05_result.csv experiment08_result.csv experiment11_result.csv experiment14_result.csv experiment17_result.csv experiment20_result.csv
experiment03_result.csv experiment06_result.csv experiment09_result.csv experiment12_result.csv experiment15_result.csv experiment18_result.csv
otknoy@debian:~/repos/leap_practice$
experiment01_result.csv の中身
otknoy@debian:~/repos/leap_practice$ head experiment/result/experiment01_result.csv
subject_name,subject_pattern,target_pattern,baseline,approach
nishikawa,circle-fast_slow_gradually,sin(slow),48.74153576293306,15579.189644603357
nishikawa,circle-fast_slow_gradually,sin(normal),42.21128169400617,17638.22486246564
nishikawa,circle-fast_slow_gradually,sin(fast),39.0736100062717,14249.857644208825
nishikawa,circle-fast_slow_gradually,sin(slow_fast),39.4852317322955,14460.881850909991
nishikawa,circle-fast_slow_gradually,sin(fast_slow),48.40087160800845,12530.166805369845
nishikawa,circle-fast_slow_gradually,sin(slow_fast_gradually),39.364991046032216,26643.219291189824
nishikawa,circle-fast_slow_gradually,sin(fast_slow_gradually),48.8663071842649,13575.850111809468
nishikawa,circle-fast_slow_gradually,sin(slow_R),51.97090903422937,22745.086930284775
nishikawa,circle-fast_slow_gradually,sin(normal_R),46.34761278077111,21283.128722974794
otknoy@debian:~/repos/leap_practice$
ベースライン手法 (正規化 + DTW) を実装した https://github.com/asdmay/leap_practice/pull/45 こっちのプルリクをマージした後に、確認 & マージしてください
やったこと
使い方
引数に被験者より得た実験データ (被験者一人分の json ファイル) を指定して
experiment/ec_exp.js
を実行して下さい。実験データ (6つの軌跡データ) と samples.json に含まれる検索対象のデータを総当りして類似度を求めます。
実行すると以下の形式の CSV ファイルを出力します。
実行例