def new_test(model, test_file_name, test_file_length): test_file = open(test_file_name) i = 0 list_ccci = [] list_ccwi = [] list_mcci = [] list_mcwi = [] for line in test_file: correct_line = line.replace("\n", "") mispelled_line = test_file.next().replace("\n", "") i = i + 1 print "[" + str(i) + "/" + str( test_file_length) + "] Checking:\n\t " + mispelled_line corrected_tweet = TweetChecker.dull_check(mispelled_line, model) print "Calulating allignment ccci..." cc_char_identity = align_char(corrected_tweet, correct_line) print "Calulating allignment ccwi..." cc_word_identity = count_correct_words(corrected_tweet, correct_line) print "Calulating allignment mcci..." mc_char_identity = align_char(mispelled_line, correct_line) print "Calulating allignment mcwi..." mc_word_identity = count_correct_words(mispelled_line, correct_line) list_ccci.append(cc_char_identity) list_ccwi.append(cc_word_identity) list_mcci.append(mc_char_identity) list_mcwi.append(mc_word_identity) mean_ccci = numpy.mean(list_ccci) mean_ccwi = numpy.mean(list_ccwi) mean_mcci = numpy.mean(list_mcci) mean_mcwi = numpy.mean(list_mcwi) first_not_account_index = 135 accounts_mean_ccci = numpy.mean(list_ccci[:first_not_account_index]) accounts_mean_ccwi = numpy.mean(list_ccwi[:first_not_account_index]) accounts_mean_mcci = numpy.mean(list_mcci[:first_not_account_index]) accounts_mean_mcwi = numpy.mean(list_mcwi[:first_not_account_index]) others_mean_ccci = numpy.mean(list_ccci[first_not_account_index:]) others_mean_ccwi = numpy.mean(list_ccwi[first_not_account_index:]) others_mean_mcci = numpy.mean(list_mcci[first_not_account_index:]) others_mean_mcwi = numpy.mean(list_mcwi[first_not_account_index:]) std_ccci = numpy.std(list_ccci) std_ccwi = numpy.std(list_ccwi) std_mcci = numpy.std(list_mcci) std_mcwi = numpy.std(list_mcwi) results_string = "" results_string += "<CORRECTED TWEET, CORRECT TWEET> CHARS IDENTITY:\n" results_string += str(list_ccci) + "\n" results_string += "\nMEAN: " + str(mean_ccci) results_string += "\nSTD: " + str(std_ccci) results_string += "\nSAME ACCOUNTS MEAN: " + str(accounts_mean_ccci) results_string += "\nOTHER ACCOUNTS MEAN: " + str(others_mean_ccci) results_string += "\n\n" results_string += "<CORRECTED TWEET, CORRECT TWEET> CORRECT WORDS:\n" results_string += str(list_ccwi) + "\n" results_string += "\nMEAN: " + str(mean_ccwi) results_string += "\nSTD: " + str(std_ccwi) results_string += "\nSAME ACCOUNTS MEAN: " + str(accounts_mean_ccwi) results_string += "\nOTHER ACCOUNTS MEAN: " + str(others_mean_ccwi) results_string += "\n\n" results_string += "<MISPELLED TWEET, CORRECT TWEET> CHARS IDENTITY:\n" results_string += str(list_mcci) + "\n" results_string += "\nMEAN: " + str(mean_mcci) results_string += "\nSTD: " + str(std_mcci) results_string += "\nSAME ACCOUNTS MEAN: " + str(accounts_mean_mcci) results_string += "\nOTHER ACCOUNTS MEAN: " + str(others_mean_mcci) results_string += "\n\n" results_string += "<MISPELLED TWEET, CORRECT TWEET> CORRECT WORDS:\n" results_string += str(list_mcwi) + "\n" results_string += "\nMEAN: " + str(mean_mcwi) results_string += "\nSTD: " + str(std_mcwi) results_string += "\nSAME ACCOUNTS MEAN: " + str(accounts_mean_mcwi) results_string += "\nOTHER ACCOUNTS MEAN: " + str(others_mean_mcwi) results_string += "\n\n" results_string += ("GAIN:\n") results_string += "\nWORD MEAN: " + str(mean_ccwi) + " - " + str( mean_mcwi) + " = " + str(mean_ccwi - mean_mcwi) results_string += "\nWORD STD: " + str(std_ccwi) + " - " + str( std_mcwi) + " = " + str(std_ccwi - std_mcwi) results_string += "\nCHARS MEAN: " + str(mean_ccci) + " - " + str( mean_mcci) + " = " + str(mean_ccci - mean_mcci) results_string += "\nCHARS STD: " + str(std_ccci) + " - " + str( std_mcci) + " = " + str(std_ccci - std_mcci) results_string += "\n\n" return results_string