if __name__ == '__main__': total_err_ran = 0.0 total_err = 0.0 total_err1 = 0.0 total_err2 = 0.0 total_err_nno = 0.0 total_err_nnced = 0.0 total_err_nnwed = 0.0 total_err_bos = 0.0 for i in range(util.TRIALS): print "Trial", i ql = util.parsePowerGrading() ran = RandomPredictor(ql) bow = ConstantBestGuessPredictor(ql) bow1 = NGramRegression(ql, 1) bow2 = NGramRegression(ql, 2) nno = NearestNeighborOverlap(ql) nnced = NearestNeighborCharEditDistance(ql) nnwed = NearestNeighborWordEditDistance(ql) bos = BagOfSynsets(ql) total_err_ran += ran.run() total_err += bow.run() total_err1 += bow1.run() total_err2 += bow2.run() total_err_nno += nno.run() total_err_nnced += nnced.run() total_err_nnwed += nnwed.run() total_err_bos += bos.run()
if __name__ == '__main__': total_err_ran = 0.0 total_err_cbg = 0.0 total_err_bop = 0.0 total_err_fop = 0.0 total_err_edit = 0.0 total_err_cep = 0.0 total_err_com = 0.0 for i in range(util.TRIALS): print "Trial", i ql = util.parsePowerGrading() ran = RandomPredictor(ql) cbg = ConstantBestGuessPredictor(ql) bop = BOPRegression(ql) fop = FOPRegression(ql) cep = CharEditDistancePredictor(ql) edit = WordEditDistancePredictor(ql) com = CombinedModel(ql, [NGramRegression, FractionOverlapPredictor, CharEditDistancePredictor], linear_model.LinearRegression) total_err_ran += ran.run() total_err_cbg += cbg.run() total_err_bop += bop.run() total_err_fop += fop.run() total_err_edit += edit.run() total_err_cep += cep.run() total_err_com += com.run() print "Average Random RMSE:",