def get_data_segment_angles(model_id, driver_id, repeat, test=False, segment_version=1, extra=((1, 1), 2)): seed = random.Random(x=driver_id + model_id) da = DataAccess() ngram_range, min_df = extra if test: set1 = list(da.get_rides_segments(driver_id, version=segment_version)) set2 = list( da.get_random_rides(settings.BIG_CHUNK_TEST * repeat, driver_id, segments=True, version=segment_version, seed=seed)) else: driver_train, driver_test = da.get_rides_split(driver_id, settings.BIG_CHUNK, segments=True, version=segment_version) other_train = list( da.get_random_rides(settings.BIG_CHUNK * repeat, driver_id, segments=True, version=segment_version, seed=seed)) other_test = list( da.get_random_rides(settings.SMALL_CHUNK, driver_id, segments=True, version=segment_version)) set1 = driver_train + other_train set2 = driver_test + other_test # create features for each (segment, angle, segment) tuple set1 = [[ '%s_%s_%s' % (d[0][i - 1], d[1][i - 1], d[0][i]) for i in xrange(1, len(d[0])) ] for d in set1] set2 = [[ '%s_%s_%s' % (d[0][i - 1], d[1][i - 1], d[0][i]) for i in xrange(1, len(d[0])) ] for d in set2] set1 = [util.get_list_string(d) for d in set1] set2 = [util.get_list_string(d) for d in set2] vectorizer = CountVectorizer(min_df=min_df, ngram_range=ngram_range) set1 = vectorizer.fit_transform(set1) set2 = vectorizer.transform(set2) return set1, set2
def get_data_segment_lengths(model_id, driver_id, repeat, test=False, segment_version=1, extra=((1,8),1)): seed = random.Random(x=driver_id+model_id) da = DataAccess() ngram_range, min_df = extra if test: set1 = list(da.get_rides_segments(driver_id, version=segment_version)) set2 = list(da.get_random_rides( settings.BIG_CHUNK_TEST * repeat, driver_id, segments=True, version=segment_version, seed=seed )) else: driver_train, driver_test = da.get_rides_split( driver_id, settings.BIG_CHUNK, segments=True, version=segment_version ) other_train = list(da.get_random_rides( settings.BIG_CHUNK * repeat, driver_id, segments=True, version=segment_version, seed=seed )) other_test = list(da.get_random_rides( settings.SMALL_CHUNK, driver_id, segments=True, version=segment_version )) set1 = driver_train + other_train set2 = driver_test + other_test # keep only lengths set1 = [d[0] for d in set1] set2 = [d[0] for d in set2] # convert to text set1 = [util.get_list_string(d) for d in set1] set2 = [util.get_list_string(d) for d in set2] vectorizer = CountVectorizer(min_df=min_df, ngram_range=ngram_range) set1 = vectorizer.fit_transform(set1) set2 = vectorizer.transform(set2) return set1, set2
def get_data_segment_lengths(model_id, driver_id, repeat, test=False, segment_version=1, extra=((1, 8), 1)): seed = random.Random(x=driver_id + model_id) da = DataAccess() ngram_range, min_df = extra if test: set1 = list(da.get_rides_segments(driver_id, version=segment_version)) set2 = list( da.get_random_rides(settings.BIG_CHUNK_TEST * repeat, driver_id, segments=True, version=segment_version, seed=seed)) else: driver_train, driver_test = da.get_rides_split(driver_id, settings.BIG_CHUNK, segments=True, version=segment_version) other_train = list( da.get_random_rides(settings.BIG_CHUNK * repeat, driver_id, segments=True, version=segment_version, seed=seed)) other_test = list( da.get_random_rides(settings.SMALL_CHUNK, driver_id, segments=True, version=segment_version)) set1 = driver_train + other_train set2 = driver_test + other_test # keep only lengths set1 = [d[0] for d in set1] set2 = [d[0] for d in set2] # convert to text set1 = [util.get_list_string(d) for d in set1] set2 = [util.get_list_string(d) for d in set2] vectorizer = CountVectorizer(min_df=min_df, ngram_range=ngram_range) set1 = vectorizer.fit_transform(set1) set2 = vectorizer.transform(set2) return set1, set2
def get_data_segment_angles_v2(model_id, driver_id, repeat, test=False, segment_version=1, extra=((1,3),1)): seed = random.Random(x=driver_id+model_id) da = DataAccess() ngram_range, min_df = extra if test: set1 = list(da.get_rides_segments(driver_id, version=segment_version)) set2 = list(da.get_random_rides( settings.BIG_CHUNK_TEST * repeat, driver_id, segments=True, version=segment_version, seed=seed )) else: driver_train, driver_test = da.get_rides_split( driver_id, settings.BIG_CHUNK, segments=True, version=segment_version ) other_train = list(da.get_random_rides( settings.BIG_CHUNK * repeat, driver_id, segments=True, version=segment_version, seed=seed )) other_test = list(da.get_random_rides( settings.SMALL_CHUNK, driver_id, segments=True, version=segment_version )) set1 = driver_train + other_train set2 = driver_test + other_test # create features for each (segment, angle, segment) tuple set1 = [['%s_%s' % (d[0][i-1], d[1][i-1]) for i in xrange(1, len(d[0]))] for d in set1] set2 = [['%s_%s' % (d[0][i-1], d[1][i-1]) for i in xrange(1, len(d[0]))] for d in set2] set1 = [util.get_list_string(d) for d in set1] set2 = [util.get_list_string(d) for d in set2] vectorizer = CountVectorizer(min_df=min_df, ngram_range=ngram_range) set1 = vectorizer.fit_transform(set1) set2 = vectorizer.transform(set2) return set1, set2