| 1 | #! /usr/bin/python |
|---|
| 2 | |
|---|
| 3 | from aubio.bench.node import * |
|---|
| 4 | from aubio.tasks import * |
|---|
| 5 | |
|---|
| 6 | class benchonset(bench): |
|---|
| 7 | |
|---|
| 8 | def dir_eval(self): |
|---|
| 9 | self.P = 100*float(self.expc-self.missed-self.merged)/(self.expc-self.missed-self.merged + self.bad+self.doubled) |
|---|
| 10 | self.R = 100*float(self.expc-self.missed-self.merged)/(self.expc-self.missed-self.merged + self.missed+self.merged) |
|---|
| 11 | if self.R < 0: self.R = 0 |
|---|
| 12 | self.F = 2* self.P*self.R / (self.P+self.R) |
|---|
| 13 | |
|---|
| 14 | self.values = [self.params.onsetmode, |
|---|
| 15 | "%2.3f" % self.params.threshold, |
|---|
| 16 | self.orig, |
|---|
| 17 | self.expc, |
|---|
| 18 | self.missed, |
|---|
| 19 | self.merged, |
|---|
| 20 | self.bad, |
|---|
| 21 | self.doubled, |
|---|
| 22 | (self.orig-self.missed-self.merged), |
|---|
| 23 | "%2.3f" % (100*float(self.orig-self.missed-self.merged)/(self.orig)), |
|---|
| 24 | "%2.3f" % (100*float(self.bad+self.doubled)/(self.orig)), |
|---|
| 25 | "%2.3f" % (100*float(self.orig-self.missed)/(self.orig)), |
|---|
| 26 | "%2.3f" % (100*float(self.bad)/(self.orig)), |
|---|
| 27 | "%2.3f" % self.P, |
|---|
| 28 | "%2.3f" % self.R, |
|---|
| 29 | "%2.3f" % self.F ] |
|---|
| 30 | |
|---|
| 31 | def file_exec(self,input,output): |
|---|
| 32 | filetask = self.task(input,params=self.params) |
|---|
| 33 | computed_data = filetask.compute_all() |
|---|
| 34 | results = filetask.eval(computed_data) |
|---|
| 35 | self.orig += filetask.orig |
|---|
| 36 | self.missed += filetask.missed |
|---|
| 37 | self.merged += filetask.merged |
|---|
| 38 | self.expc += filetask.expc |
|---|
| 39 | self.bad += filetask.bad |
|---|
| 40 | self.doubled += filetask.doubled |
|---|
| 41 | |
|---|
| 42 | |
|---|
| 43 | def run_bench(self,modes=['dual'],thresholds=[0.5]): |
|---|
| 44 | self.modes = modes |
|---|
| 45 | self.thresholds = thresholds |
|---|
| 46 | |
|---|
| 47 | self.pretty_print(self.titles) |
|---|
| 48 | for mode in self.modes: |
|---|
| 49 | self.params.onsetmode = mode |
|---|
| 50 | for threshold in self.thresholds: |
|---|
| 51 | self.params.threshold = threshold |
|---|
| 52 | self.dir_exec() |
|---|
| 53 | self.dir_eval() |
|---|
| 54 | self.pretty_print(self.values) |
|---|
| 55 | |
|---|
| 56 | def auto_learn(self,modes=['dual'],thresholds=[0.1,1.5]): |
|---|
| 57 | """ simple dichotomia like algorithm to optimise threshold """ |
|---|
| 58 | self.modes = modes |
|---|
| 59 | self.pretty_print(self.titles) |
|---|
| 60 | for mode in self.modes: |
|---|
| 61 | steps = 10 |
|---|
| 62 | lesst = thresholds[0] |
|---|
| 63 | topt = thresholds[1] |
|---|
| 64 | self.params.onsetmode = mode |
|---|
| 65 | |
|---|
| 66 | self.params.threshold = topt |
|---|
| 67 | self.dir_exec() |
|---|
| 68 | self.dir_eval() |
|---|
| 69 | self.pretty_print(self.values) |
|---|
| 70 | topF = self.F |
|---|
| 71 | |
|---|
| 72 | self.params.threshold = lesst |
|---|
| 73 | self.dir_exec() |
|---|
| 74 | self.dir_eval() |
|---|
| 75 | self.pretty_print(self.values) |
|---|
| 76 | lessF = self.F |
|---|
| 77 | |
|---|
| 78 | for i in range(steps): |
|---|
| 79 | self.params.threshold = ( lesst + topt ) * .5 |
|---|
| 80 | self.dir_exec() |
|---|
| 81 | self.dir_eval() |
|---|
| 82 | self.pretty_print(self.values) |
|---|
| 83 | if self.F == 100.0 or self.F == topF: |
|---|
| 84 | print "assuming we converged, stopping" |
|---|
| 85 | break |
|---|
| 86 | #elif abs(self.F - topF) < 0.01 : |
|---|
| 87 | # print "done converging" |
|---|
| 88 | # break |
|---|
| 89 | if topF < self.F: |
|---|
| 90 | #lessF = topF |
|---|
| 91 | #lesst = topt |
|---|
| 92 | topF = self.F |
|---|
| 93 | topt = self.params.threshold |
|---|
| 94 | elif lessF < self.F: |
|---|
| 95 | lessF = self.F |
|---|
| 96 | lesst = self.params.threshold |
|---|
| 97 | if topt == lesst: |
|---|
| 98 | lesst /= 2. |
|---|
| 99 | |
|---|
| 100 | |
|---|
| 101 | if __name__ == "__main__": |
|---|
| 102 | import sys |
|---|
| 103 | if len(sys.argv) > 1: datapath = sys.argv[1] |
|---|
| 104 | else: print "ERR: a path is required"; sys.exit(1) |
|---|
| 105 | modes = ['complex', 'energy', 'phase', 'specdiff', 'kl', 'mkl', 'dual'] |
|---|
| 106 | #modes = [ 'complex' ] |
|---|
| 107 | thresholds = [ 0.01, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5] |
|---|
| 108 | #thresholds = [1.5] |
|---|
| 109 | |
|---|
| 110 | #datapath = "%s%s" % (DATADIR,'/onset/DB/*/') |
|---|
| 111 | respath = '/var/tmp/DB-testings' |
|---|
| 112 | |
|---|
| 113 | benchonset = benchonset(datapath,respath,checkres=True,checkanno=True) |
|---|
| 114 | benchonset.params = taskparams() |
|---|
| 115 | benchonset.task = taskonset |
|---|
| 116 | |
|---|
| 117 | benchonset.titles = [ 'mode', 'thres', 'orig', 'expc', 'missd', 'mergd', |
|---|
| 118 | 'bad', 'doubl', 'corrt', 'GD', 'FP', 'GD-merged', 'FP-pruned', |
|---|
| 119 | 'prec', 'recl', 'dist' ] |
|---|
| 120 | benchonset.formats = ["%12s" , "| %6s", "| %6s", "| %6s", "| %6s", "| %6s", |
|---|
| 121 | "| %6s", "| %6s", "| %6s", "| %8s", "| %8s", "| %8s", "| %8s", |
|---|
| 122 | "| %6s", "| %6s", "| %6s"] |
|---|
| 123 | |
|---|
| 124 | try: |
|---|
| 125 | benchonset.auto_learn(modes=modes) |
|---|
| 126 | #benchonset.run_bench(modes=modes) |
|---|
| 127 | except KeyboardInterrupt: |
|---|
| 128 | sys.exit(1) |
|---|
