def evaluate_all_combinations(patfile, trainedforestbase, trainedforesttend, config): original_usertag = config.get("general", "usertag") ### remember what the tag started as original_variables = config.get("mvsc", "z") ### remember the variables print (original_variables) for tstart in ["no_tstart", "yes_tstart"]: ### iterate over all possible combinations for tend in ["no_tend", "yes_tend"]: config.set("general", "usertag", "%s-%s-%s"%(original_usertag, tstart, tend) ) ### change the usertag as needed if(tstart == "no_tstart" and tend == "no_tend"): config.set("mvsc", "z", "%s,%s,%s"%(original_variables, "time_locked", "time_end")) elif(tstart == "no_tstart"): config.set("mvsc", "z", "%s,%s"%(original_variables, "time_locked")) elif(tend == "no_tend"): config.set("mvsc", "z", "%s,%s"%(original_variables, "time_end")) else: config.set("mvsc","z", original_variables) trainedforest = trainedforestbase + "%s-%s/"%(tstart,tend) + trainedforestend ranked_file = trainedforest[:-3] + "dat" gps_start_time = -np.infty gps_end_time = np.infty dir = "." classifiersD, mla, ovl = idq.config_to_classifiersD( config ) classD = classifiersD['mvsc'] miniconfig = classD['config'] print(ranked_file) print(trainedforest) print(config.get("mvsc", "z")) idq.forest_evaluate(patfile, trainedforest, ranked_file, miniconfig, gps_start_time, gps_end_time, dir) config.set("general", "usertag", original_usertag) config.set("mvsc", "z", original_variables)
### read global configuration file config = ConfigParser.SafeConfigParser() config.read(opts.config) ifo = config.get('general', 'ifo') usertag = config.get('general', 'usertag') if usertag: usertag = "_%s"%usertag #======================== # which classifiers #======================== ### ensure we have a section for each classifier and fill out dictionary of options classifiersD, mla, ovl = idq.config_to_classifiersD( config ) classifiers = sorted(classifiersD.keys()) if mla: ### reading parameters from config file needed for mla # auxmvc_coinc_window = config.getfloat('build_auxmvc_vectors','time-window') # auxmc_gw_signif_thr = config.getfloat('build_auxmvc_vectors','signif-threshold') auxmvc_coinc_window = config.getfloat('realtime', 'padding') auxmc_gw_signif_thr = config.getfloat('general', 'gw_kwsignif_thr') auxmvc_selected_channels = config.get('general','selected-channels') auxmvc_unsafe_channels = config.get('general','unsafe-channels') #min_samples = config.getint('train', 'min_samples') ### minimum number of samples a training set should have #min_svm_samples = config.getint('idq_train', 'min_svm_samples')
config = ConfigParser.SafeConfigParser() config.read(opts.config) #mainidqdir = config.get('general', 'idqdir') ### get the main directory where idq pipeline is going to be running. ifo = config.get('general', 'ifo') usertag = config.get('general', 'usertag') if usertag: usertag = "_%s" % usertag #======================== # which classifiers #======================== ### ensure we have a section for each classifier and fill out dictionary of options classifiersD, mla, ovl = idq.config_to_classifiersD(config) ### get combiners information and add these to classifiersD combinersD, referenced_classifiers = idq.config_to_combinersD(config) for combiner, value in combinersD.items(): classifiersD[combiner] = value classifiers = sorted(classifiersD.keys()) #if mla: # ### reading parameters from config file needed for mla # auxmvc_coinc_window = config.getfloat('build_auxmvc_vectors','time-window') # auxmc_gw_signif_thr = config.getfloat('build_auxmvc_vectors','signif-threshold') # auxmvc_selected_channels = config.get('general','selected-channels') # auxmvc_unsafe_channels = config.get('general','unsafe-channels')