def get_most_listened_artists(number_of_artists=1,
db_file=HISTORY_DATABASE,
listens=None,
artists=None):
"""Returns the most listened to artist(s) as a tuple with description
and sorted pandas.Dataframe
result shape:
(description, pandas.Dataframe)
Description: The <number_of_artists> most listened to artists in the
dataset
Dataframe shape:
id_x artist_id artist_name count
int int string int
... ... ... ...
"""
description = "The {} most listened to artists in the dataset"\
.format(number_of_artists)
if listens is None or artists is None:
engine = dbt.get_connectable(db_file)
if listens is None:
listens = pd.read_sql_table("listens", con=engine)
if artists is None:
artists = pd.read_sql_table("artists", con=engine)
count = listens[["id", "artist_id"]].groupby("artist_id").count()\
.rename(index=int, columns={"id": "count"})
data = pd.merge(artists[["id", "artist_name"]], count[["count"]],
left_on="id", right_on="artist_id")\
.nlargest(number_of_artists, "count")
return (description, data)
def read_db():
engine = dbt.get_connectable(HISTORY_DATABASE)
listens = pd.read_sql_table("listens", con=engine)
tracks = pd.read_sql_table("tracks", con=engine)
albums = pd.read_sql_table("albums", con=engine)
artists = pd.read_sql_table("artists", con=engine)
return listens, tracks, albums, artists
def get_num_artists(artists=None, db_file=HISTORY_DATABASE):
"""Returns number of artists in the dataframe or database.
If artists is set db_file will be ignored.
"""
description = "Number of unique artists in dataset"
if artists is None:
engine = dbt.get_connectable(db_file)
artists = pd.read_sql_table("artists", con=engine)
num_artists = artists.id.nunique()
return (description, num_artists)
def get_num_tracks(tracks=None, db_file=HISTORY_DATABASE):
"""Returns number of tracks in the dataframe og database.
If tracks is set db_file will be ignored.
"""
description = "Number of unique tracks in dataset"
if tracks is None:
engine = dbt.get_connectable(db_file)
tracks = pd.read_sql_table("tracks", con=engine)
num_tracks = tracks.id.nunique()
return (description, num_tracks)
def get_num_listens(listens=None, db_file=HISTORY_DATABASE):
"""Returns number of listens in the dataframe or database.
If listens is set db_file will be ignored.
"""
description = "Total number of listens in dataset"
if listens is None:
engine = dbt.get_connectable(db_file)
listens = pd.read_sql_table("listens", con=engine)
num_listens = listens.id.nunique()
return (description, num_listens)