def test_get_top_artists(self): mapped_listens = utils.read_files_from_HDFS(self.mapped_listens_path) top_artist_limit = 1 test_top_artist = candidate_sets.get_top_artists( mapped_listens, top_artist_limit, []) cols = [ 'top_artist_credit_id', 'top_artist_name', 'user_name', 'total_count' ] self.assertListEqual(cols, test_top_artist.columns) self.assertEqual(test_top_artist.count(), 2) top_artist_id = sorted( [row.top_artist_credit_id for row in test_top_artist.collect()]) self.assertEqual(top_artist_id[0], 2) self.assertEqual(top_artist_id[1], 2) # empty df mapped_listens = mapped_listens.select('*').where( f.col('user_name') == 'lala') with self.assertRaises(TopArtistNotFetchedException): candidate_sets.get_top_artists(mapped_listens, top_artist_limit, []) with self.assertRaises(TopArtistNotFetchedException): candidate_sets.get_top_artists(mapped_listens, top_artist_limit, ['lala'])
def test_filter_last_x_days_recordings(self): mapped_listens_df = utils.read_files_from_HDFS( self.mapped_listens_path) mapped_listens_subset = utils.read_files_from_HDFS( self.mapped_listens_subset_path) recordings_df = create_dataframes.get_recordings_df( mapped_listens_df, {}) users = create_dataframes.get_users_dataframe(mapped_listens_df, {}) mapped_listens_subset = utils.read_files_from_HDFS( self.mapped_listens_subset_path) top_artist_limit = 1 top_artist_df = candidate_sets.get_top_artists(mapped_listens_subset, top_artist_limit, []) _, candidate_set_df = candidate_sets.get_top_artist_candidate_set( top_artist_df, recordings_df, users, mapped_listens_subset) df = candidate_sets.filter_last_x_days_recordings( candidate_set_df, mapped_listens_subset) user_name = [row.user_name for row in df.collect()] self.assertEqual(sorted(user_name), ['rob', 'rob', 'vansika_1']) received_recording_mbid = sorted( [row.mb_recording_mbid for row in df.collect()]) expected_recording_mbid = sorted([ "sf5a56f4-1f83-4681-b319-70a734d0d047", "af5a56f4-1f83-4681-b319-70a734d0d047", "sf5a56f4-1f83-4681-b319-70a734d0d047" ]) self.assertEqual(expected_recording_mbid, received_recording_mbid)
def test_get_top_artist_candidate_set(self): mapped_listens_df = utils.read_files_from_HDFS( self.mapped_listens_path) recordings_df = create_dataframes.get_recordings_df( mapped_listens_df, {}) users = create_dataframes.get_users_dataframe(mapped_listens_df, {}) mapped_listens_subset = utils.read_files_from_HDFS( self.mapped_listens_subset_path) top_artist_limit = 1 top_artist_df = candidate_sets.get_top_artists(mapped_listens_subset, top_artist_limit, []) top_artist_candidate_set_df, top_artist_candidate_set_df_html = candidate_sets.get_top_artist_candidate_set( top_artist_df, recordings_df, users, mapped_listens_subset) cols = ['recording_id', 'user_id', 'user_name'] self.assertListEqual(sorted(cols), sorted(top_artist_candidate_set_df.columns)) self.assertEqual(top_artist_candidate_set_df.count(), 3) cols = [ 'top_artist_credit_id', 'top_artist_name', 'mb_artist_credit_id', 'mb_artist_credit_mbids', 'mb_recording_mbid', 'msb_artist_credit_name_matchable', 'msb_recording_name_matchable', 'recording_id', 'user_name', 'user_id' ] self.assertListEqual(sorted(cols), sorted(top_artist_candidate_set_df_html.columns)) self.assertEqual(top_artist_candidate_set_df_html.count(), 3)
def test_get_top_artists(self): mapped_df = self.get_mapped_listens() test_top_artist_df = candidate_sets.get_top_artists(mapped_df, 'vansika') self.assertListEqual(['mb_artist_credit_id', 'artist_name', 'count'], test_top_artist_df.columns) self.assertEqual(test_top_artist_df.count(), 1) row = test_top_artist_df.collect()[0] self.assertEqual(row.mb_artist_credit_id, 1)