def with_pre_load_data_set():
logger = logging.getLogger(__name__)
song_set_df = get_song_set_df()
preference_statistic = PreferenceAnalytics(
users_preferences_df=get_users_preference_df()
)
preference_statistic.run()
for metadata, pt_graph_name in zip(METADATA_TO_PROCESS_LIST, METADATA_TO_PROCESS_LIST_PT):
metadata_to_process_list = ['id', metadata]
logger.info("*" * 60)
logger.info(
"*\tProcessando o metadado - "
+ str(metadata)
)
logger.info("*" * 60)
one_metadata_process(song_set_df=song_set_df.filter(metadata_to_process_list, axis=1),
users_preferences_df=preference_statistic.get_users_relevance_preferences_df(
user_top_n_relevance=USER_SIZE), preference_statistic=preference_statistic,
label=pt_graph_name)
one_metadata_process(song_set_df=song_set_df.filter(['id', 'album', 'title'], axis=1),
users_preferences_df=preference_statistic.get_users_relevance_preferences_df(
user_top_n_relevance=USER_SIZE), preference_statistic=preference_statistic,
label='|AL|+|TL|')
one_metadata_process(
song_set_df=concat_metadata_preserve_id(df_list=song_set_df, metadata_to_process_list=['album', 'title'],
new_column='AL+TL'),
users_preferences_df=preference_statistic.get_users_relevance_preferences_df(
user_top_n_relevance=USER_SIZE),
preference_statistic=preference_statistic,
label='AL+AR')
preference_statistic.print_song_statistical()
preference_statistic.print_user_statistical()
preference_statistic.make_graphics()
make_evaluate_graphics()
def data_analysis():
song_set_df = get_song_set_df()
preference_statistic = PreferenceAnalytics(
users_preferences_df=get_users_preference_df(song_set_df),
song_df=song_set_df)
preference_statistic.run()
preference_statistic.print_song_statistical()
preference_statistic.print_user_statistical()
preference_statistic.make_graphics()
def pre_load_data_set_and_song_variation_all_combination():
logger = logging.getLogger(__name__)
song_set_df = get_song_set_df()
preference_statistic = PreferenceAnalytics(
users_preferences_df=get_users_preference_df(song_set_df),
song_df=song_set_df)
preference_statistic.run()
for song_set_size in SONG_SET_SIZE_LIST:
song_set_with_size_df = song_select(
song_set_df, song_set_size,
preference_statistic.get_song_relevance_df())
preference_statistic_with_size = PreferenceAnalytics(
users_preferences_df=get_users_preference_df(
song_set_with_size_df),
song_df=song_set_with_size_df)
preference_statistic_with_size.run()
for metadata, pt_graph_name in zip(METADATA_TO_PROCESS_LIST,
METADATA_TO_PROCESS_LIST):
gc.collect()
metadata_to_process_list = ['id', metadata]
logger.info("*" * 60)
logger.info("*\tEXPERIMENTO 1 - " + str(metadata))
logger.info("*" * 60)
one_metadata_process(
song_set_df=song_set_with_size_df.filter(
metadata_to_process_list, axis=1),
users_preferences_df=preference_statistic_with_size.
get_users_relevance_preferences_df(
user_top_n_relevance=USER_SIZE),
preference_statistic=preference_statistic_with_size,
label=pt_graph_name)
gc.collect()
logger.info("*" * 60)
logger.info("*\tEXPERIMENTO 2 - " + "album and title - |AL|+|TL|")
logger.info("*" * 60)
one_metadata_process(
song_set_df=song_set_with_size_df.filter(['id', 'album', 'title'],
axis=1),
users_preferences_df=preference_statistic_with_size.
get_users_relevance_preferences_df(user_top_n_relevance=USER_SIZE),
preference_statistic=preference_statistic_with_size,
label='|AL|+|TL|')
logger.info("*" * 60)
logger.info("*\tEXPERIMENTO 2 - " + "album and artist - |AL|+|AR|")
logger.info("*" * 60)
one_metadata_process(
song_set_df=song_set_with_size_df.filter(['id', 'album', 'artist'],
axis=1),
users_preferences_df=preference_statistic_with_size.
get_users_relevance_preferences_df(user_top_n_relevance=USER_SIZE),
preference_statistic=preference_statistic_with_size,
label='|AL|+|AR|')
logger.info("*" * 60)
logger.info("*\tEXPERIMENTO 2 - " + "title and artist - |TL|+|AR|")
logger.info("*" * 60)
one_metadata_process(
song_set_df=song_set_with_size_df.filter(['id', 'title', 'artist'],
axis=1),
users_preferences_df=preference_statistic_with_size.
get_users_relevance_preferences_df(user_top_n_relevance=USER_SIZE),
preference_statistic=preference_statistic_with_size,
label='|TL|+|AR|')
gc.collect()
logger.info("*" * 60)
logger.info("*\tEXPERIMENTO 3 - " + "title and album - AL+TL")
logger.info("*" * 60)
one_metadata_process(
song_set_df=concat_metadata_preserve_id(
df_list=song_set_with_size_df,
metadata_to_process_list=['album', 'title'],
new_column='AL+TL'),
users_preferences_df=preference_statistic_with_size.
get_users_relevance_preferences_df(user_top_n_relevance=USER_SIZE),
preference_statistic=preference_statistic_with_size,
label='AL+TL')
logger.info("*" * 60)
logger.info("*\tEXPERIMENTO 3 - " + "artist and album - AL+AR")
logger.info("*" * 60)
one_metadata_process(
song_set_df=concat_metadata_preserve_id(
df_list=song_set_with_size_df,
metadata_to_process_list=['album', 'artist'],
new_column='AL+AR'),
users_preferences_df=preference_statistic_with_size.
get_users_relevance_preferences_df(user_top_n_relevance=USER_SIZE),
preference_statistic=preference_statistic_with_size,
label='AL+AR')
logger.info("*" * 60)
logger.info("*\tEXPERIMENTO 3 - " + "title and artist - TL+AR")
logger.info("*" * 60)
one_metadata_process(
song_set_df=concat_metadata_preserve_id(
df_list=song_set_with_size_df,
metadata_to_process_list=['title', 'artist'],
new_column='TL+AR'),
users_preferences_df=preference_statistic_with_size.
get_users_relevance_preferences_df(user_top_n_relevance=USER_SIZE),
preference_statistic=preference_statistic_with_size,
label='TL+AR')
# logger.info("*" * 60)
# logger.info(
# "*\tEXPERIMENTO 4 - "
# + "title, artist and album - TL+AR+AL"
# )
# logger.info("*" * 60)
# one_metadata_process(
# song_set_df=concat_metadata_preserve_id(df_list=song_set_with_size_df,
# metadata_to_process_list=['title', 'artist', 'album'],
# new_column='TL+AR+AL'),
# users_preferences_df=preference_statistic_with_size.get_users_relevance_preferences_df(
# user_top_n_relevance=USER_SIZE),
# preference_statistic=preference_statistic_with_size,
# label='TL+AR+AL'
# )
# gc.collect()
# logger.info("*" * 60)
# logger.info(
# "*\tEXPERIMENTO 4 - "
# + "title, artist and album- |TL|+|AR|+|AL|"
# )
# logger.info("*" * 60)
# one_metadata_process(
# song_set_df=song_set_with_size_df.filter(['id', 'title', 'artist', 'album'], axis=1),
# users_preferences_df=preference_statistic_with_size.get_users_relevance_preferences_df(
# user_top_n_relevance=USER_SIZE),
# preference_statistic=preference_statistic_with_size,
# label='|TL|+|AR|+|AL|'
# )
preference_statistic_with_size.print_song_statistical()
preference_statistic_with_size.print_user_statistical()
preference_statistic_with_size.make_graphics()
gc.collect()
make_evaluate_graphics()