def advanced_season_totals_to_csv(season):
next_season = season + 1
client.players_advanced_season_totals(
season_end_year=season,
output_type=OutputType.CSV,
output_file_path="./" + season + "_" + next_season + "_advanced_player_season_totals.csv"
)
def assert_player_advanced_season_totals_csv(self):
players_advanced_season_totals(
season_end_year=self.year,
output_type=OutputType.CSV,
output_file_path=self.output_file_path,
include_combined_values=self.include_combined_values,
)
with open(self.output_file_path, "r", encoding="utf8") as output_file, \
open(self.expected_output_file_path, "r", encoding="utf8") as expected_output_file:
self.assertEqual(
output_file.readlines(),
expected_output_file.readlines(),
)
def assert_player_advanced_season_totals_json(self):
players_advanced_season_totals(
season_end_year=self.year,
output_type=OutputType.JSON,
output_file_path=self.output_file_path,
include_combined_values=self.include_combined_values,
)
with open(self.output_file_path, "r", encoding="utf8") as output_file, \
open(self.expected_output_file_path, "r", encoding="utf8") as expected_output_file:
self.assertEqual(
json.load(output_file),
json.load(expected_output_file),
)
def test_last_2018_players_advanced_season_totals_row(self):
result = players_advanced_season_totals(season_end_year=2018)
self.assertEqual(
result[604], {
"age": 20,
"assist_percentage": 8.8,
"block_percentage": 3.0,
"box_plus_minus": -2.0,
"defensive_box_plus_minus": -0.3,
"defensive_rebound_percentage": 20.1,
"defensive_win_shares": 0.5,
"free_throw_attempt_rate": 0.418,
"games_played": 43,
"is_combined_totals": False,
"minutes_played": 410,
"name": "Ivica Zubac",
"offensive_box_plus_minus": -1.8,
"offensive_rebound_percentage": 11.8,
"offensive_win_shares": 0.5,
"player_efficiency_rating": 15.3,
"positions": [Position.CENTER],
"slug": "zubaciv01",
"steal_percentage": 0.9,
"team": Team.LOS_ANGELES_LAKERS,
"three_point_attempt_rate": 0.008,
"total_rebound_percentage": 16.0,
"true_shooting_percentage": 0.557,
"turnover_percentage": 15.3,
"usage_percentage": 17.6,
"value_over_replacement_player": 0.0,
"win_shares": 1.0,
"win_shares_per_48_minutes": 0.118
})
def player_stats():
print("collecting player season totals")
client.players_season_totals(
season_end_year=2020,
output_type=OutputType.CSV,
output_file_path=
"C:\\Users\\NWHAL\\Documents\\nba_project\\2020_player_season_totals.csv"
)
print("collecting advanced player stats")
client.players_advanced_season_totals(
season_end_year=2020,
output_type=OutputType.CSV,
output_file_path=
"C:\\Users\\NWHAL\\Documents\\nba_project\\2020_advanced_player_season_totals.csv"
)
def get_rookie_advanced_stats(year, rookie_names):
season_totals = client.players_advanced_season_totals(season_end_year=year)
stats_dict = {}
for total in season_totals:
if total['name'] in rookie_names and total['games_played'] >= 30:
stats_dict[total['name']] = total
return stats_dict
def get_roster_player_list(team):
client.players_advanced_season_totals(
season_end_year=2020,
output_type=OutputType.CSV,
output_file_path="./csv/advanced_players.csv")
client.players_season_totals(season_end_year=2020,
output_type=OutputType.CSV,
output_file_path="./csv/season_totals.csv")
df_advanced_stats = pd.read_csv("./csv/advanced_players.csv")
df_total_season_stats = pd.read_csv("./csv/season_totals.csv")
players_advanced_roster_stats = df_advanced_stats.loc[
df_advanced_stats['team'].str.contains(team.name[1].upper())]
players_total_roster_stats = df_total_season_stats.loc[
df_advanced_stats['team'].str.contains(team.name[1].upper())]
return players_advanced_roster_stats, players_total_roster_stats
def plot_usage_vs_points():
df = DataFrame()
df2 = DataFrame()
for x in (client.players_advanced_season_totals(season_end_year=2020)):
df = df.append(DataFrame(x), ignore_index=True)
for x in (client.players_season_totals(season_end_year=2020)):
df2 = df2.append(DataFrame(x), ignore_index=True)
df['avg_points'] = round(df2['points'] / df2['games_played'], 2)
df.plot(x='usage_percentage', y='avg_points', kind='scatter')
plt.show()
def test_players_advanced_season_totals_json(self):
expected_output_file_path = os.path.join(
os.path.dirname(__file__),
"../output/expected/player_advanced_season_totals_2018.json",
)
result = players_advanced_season_totals(season_end_year=2018,
output_type=OutputType.JSON)
with open(expected_output_file_path, "r") as expected_output:
self.assertEqual(
json.loads(result),
json.load(expected_output),
)
def get_season_advanced():
"""
Scrapes advanced stats for all seasons since 1950
"""
df = pd.DataFrame()
for year in range(1950, 2020):
players = client.players_advanced_season_totals(season_end_year=year)
for player in players:
player["year"] = year
df = df.append(player, ignore_index=True)
return df
def create_player_totals_csv(year_start, year_end, advanced=False):
for i in range(year_start, year_end + 1):
if advanced:
if not os.path.exists("data/adv_total_stats"):
os.makedirs("data/adv_total_stats")
client.players_advanced_season_totals(
season_end_year=i,
output_type=OutputType.CSV,
output_file_path="adv_total_stats/stats_{}.csv".format(i))
else:
if not os.path.exists("data/total_stats"):
os.makedirs("data/total_stats")
client.players_season_totals(
season_end_year=i,
output_type=OutputType.CSV,
output_file_path="total_stats/stats_{}.csv".format(i))
# self.feed_season_stats_to_db(stats, i)
print(
"Exported stats for the year {}. {} percent completed".format(
i, 100 * ((i - year_start + 1) /
(year_end - year_start + 1))))
def get_df_advanced_player_stats(year):
"""
Aggregates all relevant ADVANCED stats for players in given year
:param year: Integer Ex.) 2021
:return: DataFrame
"""
adv = client.players_advanced_season_totals(season_end_year=year)
df_advanced = pd.json_normalize(adv)
df_advanced["position"] = df_advanced["positions"].astype(str).str.split(":").str[1]
df_advanced["position"] = df_advanced["position"].map(lambda x: x.lstrip(" '").rstrip(">]'"))
cols_groupby = ["slug", "name", "age", "position"]
cols_adv = ['player_efficiency_rating',
'true_shooting_percentage', 'three_point_attempt_rate',
'free_throw_attempt_rate', 'offensive_rebound_percentage',
'defensive_rebound_percentage', 'total_rebound_percentage',
'assist_percentage', 'steal_percentage', 'block_percentage',
'turnover_percentage', 'usage_percentage', 'offensive_win_shares',
'defensive_win_shares', 'win_shares', 'win_shares_per_48_minutes',
'offensive_box_plus_minus', 'defensive_box_plus_minus',
'box_plus_minus', 'value_over_replacement_player']
cols_adv_contrib = [col + "_contrib" for col in cols_adv]
df_total_minutes = df_advanced.groupby(cols_groupby)["minutes_played"].sum().reset_index()
df_total_minutes = df_total_minutes.rename(columns = {"minutes_played": "minutues_played_total"})
df_stats = pd.merge(df_advanced, df_total_minutes, on=cols_groupby)
df_stats["weight"] = df_stats["minutes_played"] / df_stats["minutues_played_total"]
print(df_stats.columns)
# Get contribs
for col in cols_adv:
df_stats[col+"_contrib"] = df_stats[col] * df_stats["weight"]
# Aggregate contributions to get final df
df_stats_agg = df_stats.groupby(cols_groupby)[cols_adv_contrib].sum().reset_index()
# Remove "contrib" column name from all contrib columns
cols_df_stats_agg = [col.replace("_contrib", "") for col in df_stats_agg.columns]
df_stats_agg.columns = cols_df_stats_agg
df_stats.to_pickle(c.PICKLE_PATH_ADV_STATS)
return df_stats_agg
def clean_advanced(year):
advanced_stats = client.players_advanced_season_totals(
season_end_year=year)
df = pd.DataFrame(advanced_stats)
# Handle quirk in data where traded players are represented as multiple observations
df['minutes_played_total'] = df.groupby('name').minutes_played.transform(
'sum')
df['proportion'] = df['minutes_played'] / df['minutes_played_total']
num_cols = list(df.select_dtypes(include=['int', 'float64']))
unwanted_num_cols = ['age', 'minutes_played', 'games_played']
for col in unwanted_num_cols:
num_cols.remove(col)
for col in num_cols:
df[col] = df[col] * df['proportion']
df_grouped = df.groupby('name')[num_cols].agg('sum')
df_grouped['age'] = df.groupby('name')['age'].agg('mean')
df_grouped['year'] = year
return df_grouped
def test_first_2018_players_advanced_season_totals_row(self):
result = players_advanced_season_totals(season_end_year=2018)
self.assertEqual(
result[0],
{
"age": 24,
"assist_percentage": 3.4,
"block_percentage": 0.6,
"box_plus_minus": -1.5,
"defensive_box_plus_minus": 0.4,
"defensive_rebound_percentage": 8.9,
"defensive_win_shares": 1.0,
"free_throw_attempt_rate": 0.158,
"games_played": 75,
"is_combined_totals": False,
"minutes_played": 1134,
"name": "\u00c1lex Abrines",
"offensive_box_plus_minus": -1.9,
"offensive_rebound_percentage": 2.5,
"offensive_win_shares": 1.3,
"player_efficiency_rating": 9.0,
"positions": [
Position.SHOOTING_GUARD,
],
"slug": "abrinal01",
"steal_percentage": 1.7,
"team": Team.OKLAHOMA_CITY_THUNDER,
"three_point_attempt_rate": 0.759,
"total_rebound_percentage": 5.6,
"true_shooting_percentage": 0.567,
"turnover_percentage": 7.4,
"usage_percentage": 12.7,
"value_over_replacement_player": 0.1,
"win_shares": 2.2,
"win_shares_per_48_minutes": 0.094
},
)
def test_players_advanced_season_totals_csv_append(self):
players_advanced_season_totals(
season_end_year=2018,
output_type=OutputType.CSV,
output_file_path="./player_advanced_season_totals_2019.csv",
output_write_option=OutputWriteOption.APPEND)
def test_players_advanced_season_totals_csv(self):
players_advanced_season_totals(
season_end_year=2018,
output_type=OutputType.CSV,
output_file_path="./player_advanced_season_totals_2019.csv")
def test_players_advanced_season_totals_json(self):
result = players_advanced_season_totals(season_end_year=2018,
output_type=OutputType.JSON)
self.assertIsNotNone(result)
def test_players_advanced_season_totals(self):
result = players_advanced_season_totals(season_end_year=2018)
self.assertIsNotNone(result)
# -*- coding: utf-8 -*-
from basketball_reference_web_scraper import client
from basketball_reference_web_scraper.data import OutputType
import pandas as pd
seasons_df = pd.read_excel(io='years_nba.xlsx')
seasons_list = [row['seasons'] for index, row in seasons_df.iterrows()]
#slice seasons_list
seasons = [nba_season[0:4] for nba_season in seasons_list]
for nba_season in seasons:
file_path_thistime = "./" + nba_season + "_advanced_stats.csv"
client.players_advanced_season_totals(season_end_year=nba_season,
output_type=OutputType.CSV,
output_file_path=file_path_thistime)
def main():
print("\n#######################################################################")
print("#######################################################################")
print("######################### NBA Report exporter #########################")
print("#######################################################################")
print("#######################################################################\n")
while (True):
print(
"1. Players box scores by a date\
\n2. Players season statistics for a season\
\n3. Players advanced season statistics for a season\
\n4. All Team box scores by a date\
\n5. Schedule for a season\
\n6. Exit"
)
reportObject = input("\nPlease select a option: ")
# Players box scores by a date
if (reportObject == "1"):
inputDate = input("\nEnter a date (use this format 1-1-2018): ")
fileName = "all-player-box-report-" + inputDate + ".csv"
dateList = inputDate.split("-")
print("Exporting report please wait..........")
# Call Export function
client.player_box_scores(
day=dateList[0],
month=dateList[1],
year=dateList[2],
output_type=OutputType.CSV,
output_file_path="exported_files/" + fileName
)
print("Report exported at: " + abspath("exported_files/" + fileName) + "!!\n\n")
# Players season statistics for a season
elif (reportObject == "2"):
endYear = input("\nEnter season end year: ")
fileName = "all-player-season-report-" + endYear + ".csv"
print("Exporting report please wait..........")
# Call Export function
client.players_season_totals(
season_end_year=endYear,
output_type=OutputType.CSV,
output_file_path="exported_files/" + fileName
)
print("Report exported at: " + abspath("exported_files/" + fileName) + "!!\n\n")
# Players advanced season statistics for a season
elif (reportObject == "3"):
endYear = input("\nEnter season end year: ")
fileName = "all-player-advanced-season-report-" + endYear + ".csv"
print("Exporting report please wait..........")
# Call Export function
client.players_advanced_season_totals(
season_end_year=endYear,
output_type=OutputType.CSV,
output_file_path="exported_files/" + fileName
)
print("Report exported at: " + abspath("exported_files/" + fileName) + "!!\n\n")
# All Team box scores by a date
elif (reportObject == "4"):
inputDate = input("\nEnter a date (use this format 1-1-2018): ")
fileName = "all-team-report-" + inputDate + ".csv"
dateList = inputDate.split("-")
print("Exporting report please wait..........")
# Call Export function
client.team_box_scores(
day=dateList[0],
month=dateList[1],
year=dateList[2],
output_type=OutputType.CSV,
output_file_path="exported_files/" + fileName
)
print("Report exported at: " + abspath("exported_files/" + fileName) + "!!\n\n")
# Schedule for a season
elif (reportObject == "5"):
endYear = input("\nEnter season end year: ")
fileName = "season-schedule-" + endYear + ".csv"
print("Exporting report please wait..........")
# Call Export function
client.season_schedule(
season_end_year=endYear,
output_type=OutputType.CSV,
output_file_path="exported_files/" + fileName
)
print("Report exported at: " + abspath("exported_files/" + fileName) + "!!\n\n")
# Exit
elif (reportObject == "6"):
print("\n#######################################################################")
print("################################# Bye #################################")
print("#######################################################################\n")
break
# Error
else:
print("Invalid option!!\n\n")
def test_2019_player_advanced_season_totals(self):
player_season_totals = client.players_advanced_season_totals(season_end_year=2019)
self.assertIsNotNone(player_season_totals)
self.assertTrue(len(player_season_totals) > 0)
def get_advanced_season_totals(year):
return client.players_advanced_season_totals(season_end_year=year)
from basketball_reference_web_scraper import client
from basketball_reference_web_scraper.data import OutputType, Team
print(client.season_schedule(season_end_year=2021))
client.players_advanced_season_totals(
season_end_year=2018,
output_type=OutputType.CSV,
output_file_path="./2017_2018_player_season_totals.csv"
)
client.play_by_play(
home_team=Team.BOSTON_CELTICS,
year=2018, month=10, day=16,
output_type=OutputType.CSV,
output_file_path="./2018_10_06_BOS_PBP.csv"
)
client.regular_season_player_box_scores(
player_identifier="westbru01",
season_end_year=2018,
output_type=OutputType.CSV,
output_file_path="./2017_2018_russell_westbrook_regular_season_box_scores.csv"
)
def create_base_df(season_year):
from basketball_reference_web_scraper import client
import pandas as pd
import unicodedata
advanced_stats = client.players_advanced_season_totals(
season_end_year=2020)
season_stats = client.players_season_totals(season_end_year=2020)
total_df = pd.DataFrame(season_stats)
total_df['positions'], total_df['team'] = total_df['positions'].astype(
str), total_df['team'].astype(str)
positions = {
"[<Position.CENTER: 'CENTER'>]": "C",
"[<Position.SHOOTING_GUARD: 'SHOOTING GUARD'>]": "SG",
"[<Position.POWER_FORWARD: 'POWER FORWARD'>]": "PF",
"[<Position.SMALL_FORWARD: 'SMALL FORWARD'>]": "SF",
"[<Position.POINT_GUARD: 'POINT GUARD'>]": "PG",
"[<Position.GUARD: 'GUARD'>]": "SG"
}
team_sub = {"Team.": "", "_": ' '}
total_df = total_df.replace(positions).replace(team_sub, regex=True)
total_df = total_df.assign(
field_goal_percentage=(total_df['made_field_goals'] * 100 /
total_df['attempted_field_goals']).round(1),
three_point_field_goal_percentage=(
total_df['made_three_point_field_goals'] * 100 /
total_df['attempted_three_point_field_goals']).round(1),
free_throw_percentage=(total_df['made_free_throws'] * 100 /
total_df['attempted_free_throws']).round(1),
rebounds=total_df['offensive_rebounds'] +
total_df['defensive_rebounds']).fillna(0)
total_df['no_accents'] = total_df['name'].apply(
lambda x: unicodedata.normalize('NFD', x).encode(
'ascii', 'ignore').decode('UTF-8').replace(".", ""))
total_df.no_accents[total_df.no_accents ==
'Taurean Waller-Prince'] = 'Taurean Prince'
total_df = total_df.groupby(['name', 'slug', 'no_accents'],
as_index=False).agg({
'field_goal_percentage':
'mean',
'free_throw_percentage':
'mean',
'made_three_point_field_goals':
'sum',
'made_field_goals':
'sum',
'made_free_throws':
'sum',
'games_played':
'sum',
'attempted_field_goals':
'sum',
'attempted_free_throws':
'sum',
'rebounds':
'sum',
'assists':
'sum',
'blocks':
'sum',
'steals':
'sum',
'turnovers':
'sum',
'team':
'last'
}).drop_duplicates()
salaries = pd.read_csv("nba_beta_salary.csv", sep=",", engine='python')
total_df_with_salaries = total_df.join(salaries[['slug', '2019-20'
]].set_index('slug'),
on='slug').dropna()
total_df_with_salaries = total_df_with_salaries.drop('slug', axis=1)
total_df_with_salaries['ppg'] = (2 * (
total_df_with_salaries['made_field_goals'] - total_df_with_salaries['made_three_point_field_goals']) + \
3 * (total_df_with_salaries['made_three_point_field_goals']) +
total_df_with_salaries['made_free_throws']) / \
total_df_with_salaries['games_played']
return total_df_with_salaries
def get_advanced_totals(self, season_year):
advanced_totals = client.players_advanced_season_totals(season_year)
return advanced_totals
def test_2018_players_advanced_season_totals_length(self):
result = players_advanced_season_totals(season_end_year=2018)
self.assertEqual(len(result), 605)
def test_2001_players_advanced_season_totals_csv(self):
players_advanced_season_totals(
season_end_year=2001,
output_type=OutputType.CSV,
output_file_path="./player_advanced_season_totals_2001.csv",
output_write_option=OutputWriteOption.WRITE)
from basketball_reference_web_scraper import client
from basketball_reference_web_scraper.data import OutputType
import pandas as pd
# Get 2019-2020 advanced season statistics for all players
filename = "data/player_advanced_stats_2020.csv"
client.players_advanced_season_totals(season_end_year=2020,
output_type=OutputType.CSV, output_file_path="data/player_advanced_stats_2020.csv")
client.players_season_totals(season_end_year=2020,
output_type=OutputType.CSV, output_file_path="data/player_stats_2020.csv")
with open('data/player_advanced_stats_2020.csv') as advanced:
with open('data/player_stats_2020.csv') as regular:
newdata = open('data/player_combined_data_2020.csv', 'w+')
i = 0
reglines = regular.readlines()
for line in advanced.readlines():
newdata.write(line.replace('\n','') + reglines[i])
i += 1
# season_end_year can only be 2000 - 2020
# Visit https://github.com/jaebradley/basketball_reference_web_scraper for more info
import json
from basketball_reference_web_scraper import client
from basketball_reference_web_scraper.data import OutputType
season_end_year = 2020
total_season_stats = client.players_season_totals(season_end_year)
advanced_stats = client.players_advanced_season_totals(season_end_year)
for a, b in zip(total_season_stats, advanced_stats):
a["advanced_stats"] = b
a["positions"] = a["positions"][0].value
a["team"] = a["team"].value
del a["advanced_stats"]['positions']
del a["advanced_stats"]['team']
season_stats = json.dumps(total_season_stats)
with open('./{year}_season_totals.json'.format(year=season_end_year),
'w') as outfile:
json.dump(total_season_stats, outfile)
