import pandas as pd
import re
from datetime import timedelta
from pyquery import PyQuery as pq
from urllib.error import HTTPError
from .. import utils
from ..constants import AWAY, HOME
from ..decorators import int_property_decorator
from .constants import (BOXSCORE_ELEMENT_INDEX,
BOXSCORE_ELEMENT_SUB_INDEX,
BOXSCORE_SCHEME,
BOXSCORE_URL,
BOXSCORES_URL)
from .player import (AbstractPlayer,
_float_property_decorator,
_int_property_decorator)
from functools import wraps
[docs]def ncaaf_int_property_sub_index(func):
# Decorator dedicated to properties with sub-indices, such as pass yards
# which is indexed within a table cell but also has multiple other values
# in that same cell that need to be ignored.
@property
@wraps(func)
def wrapper(*args):
value = func(*args)
# Equivalent to the calling property's method name
field = func.__name__
try:
field_items = value.replace('--', '-').split('-')
except AttributeError:
return None
try:
return int(field_items[BOXSCORE_ELEMENT_SUB_INDEX[field]])
except (TypeError, ValueError, IndexError):
return None
return wrapper
[docs]class BoxscorePlayer(AbstractPlayer):
"""
Get player stats for an individual game.
Given a player ID, such as 'david-blough-1' for David Blough, their full
name, and all associated stats from the Boxscore page in HTML format, parse
the HTML and extract only the relevant stats for the specified player and
assign them to readable properties.
This class inherits the ``AbstractPlayer`` class. As a result, all
properties associated with ``AbstractPlayer`` can also be read directly
from this class.
As this class is instantiated from within the Boxscore class, it should not
be called directly and should instead be queried using the appropriate
players properties from the Boxscore class.
Parameters
----------
player_id : string
A player's ID according to sports-reference.com, such as
'david-blough-1' for David Blough. The player ID can be found by
navigating to the player's stats page and getting the string between
the final slash and the '.html' in the URL. In general, the ID is in
the format 'first-last-n' where 'first' is the player's first name,
'last' is the player's last name, and 'n' is a number starting at 1 for
the first time that player ID has been used and increments by 1 for
every successive player.
player_name : string
A string representing the player's first and last name, such as 'David
Blough'.
player_data : string
A string representation of the player's HTML data from the Boxscore
page. If the player appears in multiple tables, all of their
information will appear in one single string concatenated together.
"""
def __init__(self, player_id, player_name, player_data):
self._index = 0
self._player_id = player_id
self._pass_yards_per_attempt = None
self._kickoff_returns = None
self._kickoff_return_yards = None
self._average_kickoff_return_yards = None
self._punt_returns = None
self._punt_return_yards = None
self._average_punt_return_yards = None
self._extra_points_attempted = None
self._extra_point_percentage = None
self._field_goals_attempted = None
self._field_goal_percentage = None
self._points_kicking = None
self._punts = None
self._punting_yards = None
self._punting_yards_per_attempt = None
AbstractPlayer.__init__(self, player_id, player_name, player_data)
@property
def dataframe(self):
"""
Returns a ``pandas DataFrame`` containing all other relevant class
properties and value for the specified game.
"""
fields_to_include = {
'completed_passes': self.completed_passes,
'pass_attempts': self.pass_attempts,
'passing_completion': self.passing_completion,
'passing_yards': self.passing_yards,
'pass_yards_per_attempt': self.pass_yards_per_attempt,
'adjusted_yards_per_attempt': self.adjusted_yards_per_attempt,
'passing_touchdowns': self.passing_touchdowns,
'interceptions_thrown': self.interceptions_thrown,
'quarterback_rating': self.quarterback_rating,
'rush_attempts': self.rush_attempts,
'rush_yards': self.rush_yards,
'rush_yards_per_attempt': self.rush_yards_per_attempt,
'rush_touchdowns': self.rush_touchdowns,
'receptions': self.receptions,
'receiving_yards': self.receiving_yards,
'receiving_yards_per_reception':
self.receiving_yards_per_reception,
'receiving_touchdowns': self.receiving_touchdowns,
'plays_from_scrimmage': self.plays_from_scrimmage,
'yards_from_scrimmage': self.yards_from_scrimmage,
'yards_from_scrimmage_per_play':
self.yards_from_scrimmage_per_play,
'rushing_and_receiving_touchdowns':
self.rushing_and_receiving_touchdowns,
'solo_tackles': self.solo_tackles,
'assists_on_tackles': self.assists_on_tackles,
'total_tackles': self.total_tackles,
'tackles_for_loss': self.tackles_for_loss,
'sacks': self.sacks,
'interceptions': self.interceptions,
'yards_returned_from_interceptions':
self.yards_returned_from_interceptions,
'yards_returned_per_interception':
self.yards_returned_per_interception,
'interceptions_returned_for_touchdown':
self.interceptions_returned_for_touchdown,
'passes_defended': self.passes_defended,
'fumbles_recovered': self.fumbles_recovered,
'yards_recovered_from_fumble': self.yards_recovered_from_fumble,
'fumbles_recovered_for_touchdown':
self.fumbles_recovered_for_touchdown,
'fumbles_forced': self.fumbles_forced,
'kickoff_returns': self.kickoff_returns,
'kickoff_return_yards': self.kickoff_return_yards,
'average_kickoff_return_yards': self.average_kickoff_return_yards,
'kickoff_return_touchdowns': self.kickoff_return_touchdowns,
'punt_returns': self.punt_returns,
'punt_return_yards': self.punt_return_yards,
'average_punt_return_yards': self.average_punt_return_yards,
'punt_return_touchdowns': self.punt_return_touchdowns,
'extra_points_made': self.extra_points_made,
'extra_points_attempted': self.extra_points_attempted,
'extra_point_percentage': self.extra_point_percentage,
'field_goals_made': self.field_goals_made,
'field_goals_attempted': self.field_goals_attempted,
'field_goal_percentage': self.field_goal_percentage,
'points_kicking': self.points_kicking,
'punts': self.punts,
'punting_yards': self.punting_yards,
'punting_yards_per_punt': self.punting_yards_per_attempt
}
return pd.DataFrame([fields_to_include], index=[self._player_id])
@_float_property_decorator
def pass_yards_per_attempt(self):
"""
Returns a ``float`` of the average number of yards the player gained
per pass attempt.
"""
return self._pass_yards_per_attempt
@_int_property_decorator
def kickoff_returns(self):
"""
Returns an ``int`` of the number of kickoffs the player attempted to
return.
"""
return self._kickoff_returns
@_int_property_decorator
def kickoff_return_yards(self):
"""
Returns an ``int`` of the total number of yards the player gained while
the player attempted to return a kickoff.
"""
return self._kickoff_return_yards
@_float_property_decorator
def average_kickoff_return_yards(self):
"""
Returns a ``float`` of the average number of yards the player gained
per attempted kickoff return.
"""
return self._average_kickoff_return_yards
@_int_property_decorator
def punt_returns(self):
"""
Returns an ``int`` of the number of punts the player attempted to
return.
"""
return self._punt_returns
@_int_property_decorator
def punt_return_yards(self):
"""
Returns an ``int`` of the total number of yards the player gained while
the player attempted to return a punt.
"""
return self._punt_return_yards
@_float_property_decorator
def average_punt_return_yards(self):
"""
Returns a ``float`` of the average number of yards the player gained
per attempted punt return.
"""
return self._average_punt_return_yards
@_int_property_decorator
def extra_points_attempted(self):
"""
Returns an ``int`` of the total number of extra points the player
attempted.
"""
return self._extra_points_attempted
@_float_property_decorator
def extra_point_percentage(self):
"""
Returns a ``float`` of the percentage of attempted extra points the
player made. Percentage ranges from 0-100.
"""
return self._extra_point_percentage
@_int_property_decorator
def field_goals_attempted(self):
"""
Returns an ``int`` of the total number of field goals the player
attempted.
"""
return self._field_goals_attempted
@_float_property_decorator
def field_goal_percentage(self):
"""
Returns a ``float`` of the percentage of attempted field goals the
player made. Percentage ranges from 0-100.
"""
return self._field_goal_percentage
@_int_property_decorator
def points_kicking(self):
"""
Returns an ``int`` of the total number of points the player gained from
kicking field goals or extra points.
"""
return self._points_kicking
@_int_property_decorator
def punts(self):
"""
Returns an ``int`` of the number of times the player punted the ball.
"""
return self._punts
@_int_property_decorator
def punting_yards(self):
"""
Returns an ``int`` of the total number of yards the player punted the
ball.
"""
return self._punting_yards
@_float_property_decorator
def punting_yards_per_attempt(self):
"""
Returns a ``float`` of the average number of yards the player punted
per attempt.
"""
return self._punting_yards_per_attempt
[docs]class Boxscore:
"""
Detailed information about the final statistics for a game.
Stores all relevant information for a game such as the date, time,
location, result, and more advanced metrics such as the number of fumbles
from sacks, a team's passing completion, rushing touchdowns and much more.
Parameters
----------
uri : string
The relative link to the boxscore HTML page, such as
'2018-01-08-georgia'.
"""
def __init__(self, uri):
self._uri = uri
self._date = None
self._time = None
self._stadium = None
self._away_name = None
self._home_name = None
self._winner = None
self._winning_name = None
self._winning_abbr = None
self._losing_name = None
self._losing_abbr = None
self._summary = None
self._away_points = None
self._away_first_downs = None
self._away_rush_attempts = None
self._away_rush_yards = None
self._away_rush_touchdowns = None
self._away_pass_completions = None
self._away_pass_attempts = None
self._away_pass_yards = None
self._away_pass_touchdowns = None
self._away_interceptions = None
self._away_total_yards = None
self._away_fumbles = None
self._away_fumbles_lost = None
self._away_turnovers = None
self._away_penalties = None
self._away_yards_from_penalties = None
self._home_points = None
self._home_first_downs = None
self._home_rush_attempts = None
self._home_rush_yards = None
self._home_rush_touchdowns = None
self._home_pass_completions = None
self._home_pass_attempts = None
self._home_pass_yards = None
self._home_pass_touchdowns = None
self._home_interceptions = None
self._home_total_yards = None
self._home_fumbles = None
self._home_fumbles_lost = None
self._home_turnovers = None
self._home_penalties = None
self._home_yards_from_penalties = None
self._parse_game_data(uri)
def __str__(self):
"""
Return the string representation of the class.
"""
return (f'Boxscore for {self._away_name.text()} at '
f'{self._home_name.text()} ({self.date})')
def __repr__(self):
"""
Return the string representation of the class.
"""
return self.__str__()
def _retrieve_html_page(self, uri):
"""
Download the requested HTML page.
Given a relative link, download the requested page and strip it of all
comment tags before returning a pyquery object which will be used to
parse the data.
Parameters
----------
uri : string
The relative link to the boxscore HTML page, such as
'2018-01-08-georgia'.
Returns
-------
PyQuery object
The requested page is returned as a queriable PyQuery object with
the comment tags removed.
"""
url = BOXSCORE_URL % uri
try:
url_data = pq(url)
except HTTPError:
return None
return pq(utils._remove_html_comment_tags(url_data))
def _parse_game_date_and_location(self, boxscore):
"""
Retrieve the game's date and location.
The date and location of the game follow a more complicated parsing
scheme and should be handled differently from other tags. Both fields
are separated by a newline character ('\n') with the first line being
the date and the second being the location.
Parameters
----------
boxscore : PyQuery object
A PyQuery object containing all of the HTML data from the boxscore.
"""
scheme = BOXSCORE_SCHEME['time']
items = [i.text() for i in boxscore(scheme).items()]
game_info = items[0].split('\n')
time = ''
date = ''
stadium = ''
for line in game_info:
time_match = re.findall(r'(\d:\d\d|\d\d:\d\d)', line.lower())
if len(time_match) > 0:
time = line
for day in ['monday', 'tuesday', 'wednesday', 'thursday', 'friday',
'saturday', 'sunday']:
if day in line.lower():
date = line
# In general, locations are in the format 'Stadium Name - City,
# State'. Since the ' - ' characters seem to be unique to the
# location line, it should be safe to use this as a matcher.
if ' - ' in line:
stadium = line
setattr(self, '_time', time)
setattr(self, '_date', date)
setattr(self, '_stadium', stadium)
def _parse_name(self, field, boxscore):
"""
Retrieve the team's complete name tag.
Both the team's full name (embedded in the tag's text) and the team's
abbreviation are stored in the name tag which can be used to parse
the winning and losing team's information.
Parameters
----------
field : string
The name of the attribute to parse
boxscore : PyQuery object
A PyQuery object containing all of the HTML data from the boxscore.
Returns
-------
PyQuery object
The complete text for the requested tag.
"""
scheme = BOXSCORE_SCHEME[field]
return boxscore(scheme)
def _parse_summary(self, boxscore):
"""
Find the game summary including scores in each quarter.
The game summary provides further information on the points scored
during each quarter, including the final score and any overtimes if
applicable. The final output will be in a dictionary with two keys,
'away' and 'home'. The value of each key will be a list for each
respective team's score by order of the quarter, with the first element
belonging to the first quarter, similar to the following:
{
'away': [0, 7, 3, 14],
'home': [7, 7, 3, 0]
}
Parameters
----------
boxscore : PyQuery object
A PyQuery object containing all of the HTML from the boxscore.
Returns
-------
dict
Returns a ``dictionary`` representing the score for each team in
each quarter of the game.
"""
team = ['away', 'home']
summary = {'away': [], 'home': []}
game_summary = boxscore(BOXSCORE_SCHEME['summary'])
for ind, team_info in enumerate(game_summary('tbody tr').items()):
# Only pull the first N-1 items as the last element is the final
# score for each team which is already stored in an attribute, and
# shouldn't be duplicated.
for quarter in list(team_info('td[class="center"]').items())[:-1]:
# The first element contains the logo and name of the teams,
# but not any score information, and should be skipped.
if quarter('div'):
continue
try:
summary[team[ind]].append(int(quarter.text()))
except ValueError:
summary[team[ind]].append(None)
return summary
def _find_boxscore_tables(self, boxscore):
"""
Find all tables with boxscore information on the page.
Iterate through all tables on the page and see if any of them are
boxscore pages by checking if the ID is prefixed with 'box_'. If so,
add it to a list and return the final list at the end.
Parameters
----------
boxscore : PyQuery object
A PyQuery object containing all of the HTML data from the boxscore.
Returns
-------
list
Returns a ``list`` of the PyQuery objects where each object
represents a boxscore table.
"""
tables = []
valid_tables = ['passing', 'rushing_and_receiving', 'defense',
'returns', 'kicking_and_punting']
for table in boxscore('table').items():
if table.attr['id'] in valid_tables:
tables.append(table)
return tables
def _find_player_id(self, row):
"""
Find the player's ID.
Find the player's ID as embedded in the 'data-append-csv' attribute,
such as 'david-blough-1' for David Blough.
Parameters
----------
row : PyQuery object
A PyQuery object representing a single row in a boxscore table for
a single player.
Returns
-------
str
Returns a ``string`` of the player's ID, such as 'david-blough-1'
for David Blough.
"""
return row('th').attr('data-append-csv')
def _find_player_name(self, row):
"""
Find the player's full name.
Find the player's full name, such as 'David Blough'. The name is the
text displayed for a link to the player's individual stats.
Parameters
----------
row : PyQuery object
A PyQuery object representing a single row in a boxscore table for
a single player.
Returns
-------
str
Returns a ``string`` of the player's full name, such as 'David
Blough'.
"""
return row('a:first').text()
def _find_home_or_away(self, row):
"""
Determine whether the player is on the home or away team.
Next to every player is their school's name. This name can be matched
with the previously parsed home team's name to determine if the player
is a member of the home or away team.
Parameters
----------
row : PyQuery object
A PyQuery object representing a single row in a boxscore table for
a single player.
Returns
-------
str
Returns a ``string`` constant denoting whether the team plays for
the home or away team.
"""
name = row('a:last').text()
if name == self._home_name.text():
return HOME
else:
return AWAY
def _extract_player_stats(self, table, player_dict):
"""
Combine all player stats into a single object.
Since each player generally has a couple of rows worth of stats
(rushing, passing, defense, and more) on the boxscore page, both rows
should be combined into a single string object to easily query all
fields from a single object instead of determining which row to pull
metrics from.
Parameters
----------
table : PyQuery object
A PyQuery object of a single boxscore table, such as the home
team's advanced stats or the away team's basic stats.
player_dict : dictionary
A dictionary where each key is a string of the player's ID and each
value is a dictionary where the values contain the player's name,
HTML data, and a string constant indicating which team the player
is a member of.
Returns
-------
dictionary
Returns a ``dictionary`` where each key is a string of the player's
ID and each value is a dictionary where the values contain the
player's name, HTML data, and a string constant indicating which
team the player is a member of.
"""
for row in table('tbody tr').items():
player_id = self._find_player_id(row)
# Occurs when a header row is identified instead of a player.
if not player_id:
continue
name = self._find_player_name(row)
home_or_away = self._find_home_or_away(row)
try:
player_dict[player_id]['data'] += str(row).strip()
except KeyError:
player_dict[player_id] = {
'name': name,
'data': str(row).strip(),
'team': home_or_away
}
return player_dict
def _instantiate_players(self, player_dict):
"""
Create a list of player instances for both the home and away teams.
For every player listed on the boxscores page, create an instance of
the BoxscorePlayer class for that player and add them to a list of
players for their respective team.
Parameters
----------
player_dict : dictionary
A dictionary containing information for every player on the
boxscores page. Each key is a string containing the player's ID
and each value is a dictionary with the player's full name, a
string representation of their HTML stats, and a string constant
denoting which team they play for as the values.
Returns
-------
tuple
Returns a ``tuple`` in the format (away_players, home_players)
where each element is a list of player instances for the away and
home teams, respectively.
"""
home_players = []
away_players = []
for player_id, details in player_dict.items():
player = BoxscorePlayer(player_id,
details['name'],
details['data'])
if details['team'] == HOME:
home_players.append(player)
else:
away_players.append(player)
return away_players, home_players
def _find_players(self, boxscore):
"""
Find all players for each team.
Iterate through every player for both teams as found in the boxscore
tables and create a list of instances of the BoxscorePlayer class for
each player. Return lists of player instances comprising the away and
home team players, respectively.
Parameters
----------
boxscore : PyQuery object
A PyQuery object containing all of the HTML data from the boxscore.
Returns
-------
tuple
Returns a ``tuple`` in the format (away_players, home_players)
where each element is a list of player instances for the away and
home teams, respectively.
"""
player_dict = {}
tables = self._find_boxscore_tables(boxscore)
for table in tables:
player_dict = self._extract_player_stats(table, player_dict)
away_players, home_players = self._instantiate_players(player_dict)
return away_players, home_players
def _parse_game_data(self, uri):
"""
Parses a value for every attribute.
This function looks through every attribute and retrieves the value
according to the parsing scheme and index of the attribute from the
passed HTML data. Once the value is retrieved, the attribute's value is
updated with the returned result.
Note that this method is called directly once Boxscore is invoked and
does not need to be called manually.
Parameters
----------
uri : string
The relative link to the boxscore HTML page, such as
'2018-01-08-georgia'.
"""
boxscore = self._retrieve_html_page(uri)
# If the boxscore is None, the game likely hasn't been played yet and
# no information can be gathered. As there is nothing to grab, the
# class instance should just be empty.
if not boxscore:
return
for field in self.__dict__:
# Remove the '_' from the name
short_field = str(field)[1:]
if short_field == 'winner' or \
short_field == 'winning_name' or \
short_field == 'winning_abbr' or \
short_field == 'losing_name' or \
short_field == 'losing_abbr' or \
short_field == 'uri' or \
short_field == 'date' or \
short_field == 'time' or \
short_field == 'stadium':
continue
if short_field == 'away_name' or \
short_field == 'home_name':
value = self._parse_name(short_field, boxscore)
setattr(self, field, value)
continue
if short_field == 'summary':
value = self._parse_summary(boxscore)
setattr(self, field, value)
continue
index = 0
if short_field in BOXSCORE_ELEMENT_INDEX.keys():
index = BOXSCORE_ELEMENT_INDEX[short_field]
value = utils._parse_field(BOXSCORE_SCHEME,
boxscore,
short_field,
index)
setattr(self, field, value)
self._parse_game_date_and_location(boxscore)
self._away_players, self._home_players = self._find_players(boxscore)
@property
def dataframe(self):
"""
Returns a pandas DataFrame containing all other class properties and
values. The index for the DataFrame is the string URI that is used to
instantiate the class, such as '2018-01-08-georgia'.
"""
for points in [self._away_points, self._home_points]:
if points is None or points == '':
return None
fields_to_include = {
'away_first_downs': self.away_first_downs,
'away_fumbles': self.away_fumbles,
'away_fumbles_lost': self.away_fumbles_lost,
'away_interceptions': self.away_interceptions,
'away_pass_attempts': self.away_pass_attempts,
'away_pass_completions': self.away_pass_completions,
'away_pass_touchdowns': self.away_pass_touchdowns,
'away_pass_yards': self.away_pass_yards,
'away_penalties': self.away_penalties,
'away_points': self.away_points,
'away_rush_attempts': self.away_rush_attempts,
'away_rush_touchdowns': self.away_rush_touchdowns,
'away_rush_yards': self.away_rush_yards,
'away_total_yards': self.away_total_yards,
'away_turnovers': self.away_turnovers,
'away_yards_from_penalties': self.away_yards_from_penalties,
'date': self.date,
'home_first_downs': self.home_first_downs,
'home_fumbles': self.home_fumbles,
'home_fumbles_lost': self.home_fumbles_lost,
'home_interceptions': self.home_interceptions,
'home_pass_attempts': self.home_pass_attempts,
'home_pass_completions': self.home_pass_completions,
'home_pass_touchdowns': self.home_pass_touchdowns,
'home_pass_yards': self.home_pass_yards,
'home_penalties': self.home_penalties,
'home_points': self.home_points,
'home_rush_attempts': self.home_rush_attempts,
'home_rush_touchdowns': self.home_rush_touchdowns,
'home_rush_yards': self.home_rush_yards,
'home_total_yards': self.home_total_yards,
'home_turnovers': self.home_turnovers,
'home_yards_from_penalties': self.home_yards_from_penalties,
'losing_abbr': self.losing_abbr,
'losing_name': self.losing_name,
'stadium': self.stadium,
'time': self.time,
'winner': self.winner,
'winning_abbr': self.winning_abbr,
'winning_name': self.winning_name
}
return pd.DataFrame([fields_to_include], index=[self._uri])
@property
def away_players(self):
"""
Returns a ``list`` of ``BoxscorePlayer`` class instances for each
player on the away team.
"""
return self._away_players
@property
def home_players(self):
"""
Returns a ``list`` of ``BoxscorePlayer`` class instances for each
player on the home team.
"""
return self._home_players
@property
def date(self):
"""
Returns a ``string`` of the date the game took place.
"""
return self._date
@property
def time(self):
"""
Returns a ``string`` of the time the game started.
"""
return self._time.replace('Start Time: ', '')
@property
def stadium(self):
"""
Returns a ``string`` of the name of the stadium where the game was
played.
"""
return self._stadium.replace('Stadium: ', '')
@property
def summary(self):
"""
Returns a ``dictionary`` with two keys, 'away' and 'home'. The value of
each key will be a list for each respective team's score by order of
the quarter, with the first element belonging to the first quarter,
similar to the following:
{
'away': [0, 7, 3, 14],
'home': [7, 7, 3, 0]
}
"""
return self._summary
@property
def winner(self):
"""
Returns a ``string`` constant indicating whether the home or away team
won.
"""
if self.home_points > self.away_points:
return HOME
return AWAY
@property
def winning_name(self):
"""
Returns a ``string`` of the winning team's name, such as 'Alabama'.
"""
if self.winner == HOME:
return self._home_name.text()
return self._away_name.text()
@property
def winning_abbr(self):
"""
Returns a ``string`` of the winning team's abbreviation, such as
'ALABAMA'
for the Alabama Crimson Tide.
"""
if self.winner == HOME:
if 'cfb/schools' not in str(self._home_name):
return self._home_name.text()
return utils._parse_abbreviation(self._home_name)
if 'cfb/schools' not in str(self._away_name):
return self._away_name.text()
return utils._parse_abbreviation(self._away_name)
@property
def losing_name(self):
"""
Returns a ``string`` of the losing team's name, such as 'Georgia'.
"""
if self.winner == HOME:
return self._away_name.text()
return self._home_name.text()
@property
def losing_abbr(self):
"""
Returns a ``string`` of the losing team's abbreviation, such as
'GEORGIA' for the Georgia Bulldogs.
"""
if self.winner == HOME:
if 'cfb/schools' not in str(self._away_name):
return self._away_name.text()
return utils._parse_abbreviation(self._away_name)
if 'cfb/schools' not in str(self._home_name):
return self._home_name.text()
return utils._parse_abbreviation(self._home_name)
@int_property_decorator
def away_points(self):
"""
Returns an ``int`` of the number of points the away team scored.
"""
return self._away_points
@int_property_decorator
def away_first_downs(self):
"""
Returns an ``int`` of the number of first downs the away team gained.
"""
return self._away_first_downs
@ncaaf_int_property_sub_index
def away_rush_attempts(self):
"""
Returns an ``int`` of the number of rushing plays the away team made.
"""
return self._away_rush_attempts
@ncaaf_int_property_sub_index
def away_rush_yards(self):
"""
Returns an ``int`` of the number of rushing yards the away team gained.
"""
return self._away_rush_yards
@ncaaf_int_property_sub_index
def away_rush_touchdowns(self):
"""
Returns an ``int`` of the number of rushing touchdowns the away team
scored.
"""
return self._away_rush_touchdowns
@ncaaf_int_property_sub_index
def away_pass_completions(self):
"""
Returns an ``int`` of the number of completed passes the away team
made.
"""
return self._away_pass_completions
@ncaaf_int_property_sub_index
def away_pass_attempts(self):
"""
Returns an ``int`` of the number of passes that were thrown by the away
team.
"""
return self._away_pass_attempts
@ncaaf_int_property_sub_index
def away_pass_yards(self):
"""
Returns an ``int`` of the number of passing yards the away team gained.
"""
return self._away_pass_yards
@ncaaf_int_property_sub_index
def away_pass_touchdowns(self):
"""
Returns an ``int`` of the number of passing touchdowns the away team
scored.
"""
return self._away_pass_touchdowns
@ncaaf_int_property_sub_index
def away_interceptions(self):
"""
Returns an ``int`` of the number of interceptions the away team threw.
"""
return self._away_interceptions
@int_property_decorator
def away_total_yards(self):
"""
Returns an ``int`` of the total number of yards the away team gained.
"""
return self._away_total_yards
@ncaaf_int_property_sub_index
def away_fumbles(self):
"""
Returns an ``int`` of the number of times the away team fumbled the
ball.
"""
return self._away_fumbles
@ncaaf_int_property_sub_index
def away_fumbles_lost(self):
"""
Returns an ``int`` of the number of times the away team turned the ball
over as the result of a fumble.
"""
return self._away_fumbles
@int_property_decorator
def away_turnovers(self):
"""
Returns an ``int`` of the number of times the away team turned the ball
over.
"""
return self._away_turnovers
@ncaaf_int_property_sub_index
def away_penalties(self):
"""
Returns an ``int`` of the number of penalties called on the away team.
"""
return self._away_penalties
@ncaaf_int_property_sub_index
def away_yards_from_penalties(self):
"""
Returns an ``int`` of the number of yards gifted as a result of
penalties called on the away team.
"""
return self._away_yards_from_penalties
@int_property_decorator
def home_points(self):
"""
Returns an ``int`` of the number of points the home team scored.
"""
return self._home_points
@int_property_decorator
def home_first_downs(self):
"""
Returns an ``int`` of the number of first downs the home team gained.
"""
return self._home_first_downs
@ncaaf_int_property_sub_index
def home_rush_attempts(self):
"""
Returns an ``int`` of the number of rushing plays the home team made.
"""
return self._home_rush_attempts
@ncaaf_int_property_sub_index
def home_rush_yards(self):
"""
Returns an ``int`` of the number of rushing yards the home team gained.
"""
return self._home_rush_yards
@ncaaf_int_property_sub_index
def home_rush_touchdowns(self):
"""
Returns an ``int`` of the number of rushing touchdowns the home team
scored.
"""
return self._home_rush_touchdowns
@ncaaf_int_property_sub_index
def home_pass_completions(self):
"""
Returns an ``int`` of the number of completed passes the home team
made.
"""
return self._home_pass_completions
@ncaaf_int_property_sub_index
def home_pass_attempts(self):
"""
Returns an ``int`` of the number of passes that were thrown by the home
team.
"""
return self._home_pass_attempts
@ncaaf_int_property_sub_index
def home_pass_yards(self):
"""
Returns an ``int`` of the number of passing yards the home team gained.
"""
return self._home_pass_yards
@ncaaf_int_property_sub_index
def home_pass_touchdowns(self):
"""
Returns an ``int`` of the number of passing touchdowns the home team
scored.
"""
return self._home_pass_touchdowns
@ncaaf_int_property_sub_index
def home_interceptions(self):
"""
Returns an ``int`` of the number of interceptions the home team threw.
"""
return self._home_pass_touchdowns
@int_property_decorator
def home_total_yards(self):
"""
Returns an ``int`` of the total number of yards the home team gained.
"""
return self._home_total_yards
@ncaaf_int_property_sub_index
def home_fumbles(self):
"""
Returns an ``int`` of the number of times the home team fumbled the
ball.
"""
return self._home_fumbles
@ncaaf_int_property_sub_index
def home_fumbles_lost(self):
"""
Returns an ``int`` of the number of times the home team turned the ball
over as the result of a fumble.
"""
return self._home_fumbles_lost
@int_property_decorator
def home_turnovers(self):
"""
Returns an ``int`` of the number of times the home team turned the ball
over.
"""
return self._home_turnovers
@ncaaf_int_property_sub_index
def home_penalties(self):
"""
Returns an ``int`` of the number of penalties called on the home team.
"""
return self._home_penalties
@ncaaf_int_property_sub_index
def home_yards_from_penalties(self):
"""
Returns an ``int`` of the number of yards gifted as a result of
penalties called on the home team.
"""
return self._home_yards_from_penalties
[docs]class Boxscores:
"""
Search for NCAAF games taking place on a particular day.
Retrieve a dictionary which contains a list of all games being played on a
particular day. Output includes a link to the boxscore, a boolean value
which indicates if the game is between two Division-I teams or not, and the
names and abbreviations for both the home teams. If no games are played on
a particular day, the list will be empty.
Parameters
----------
date : datetime object
The date to search for any matches. The month, day, and year are
required for the search, but time is not factored into the search.
end_date : datetime object
Optionally specify an end date to iterate until. All boxscores starting
from the date specified in the 'date' parameter up to and including the
boxscores specified in the 'end_date' parameter will be pulled. If left
empty, or if 'end_date' is prior to 'date', only the games from the day
specified in the 'date' parameter will be saved.
"""
def __init__(self, date, end_date=None):
self._boxscores = {}
self._find_games(date, end_date)
def __str__(self):
"""
Return the string representation of the class.
"""
return f"NCAAF games for {', '.join(self._boxscores.keys())}"
def __repr__(self):
"""
Return the string representation of the class.
"""
return self.__str__()
@property
def games(self):
"""
Returns a ``dictionary`` object representing all of the games played on
the requested day. Dictionary is in the following format::
{'date' : [ # 'date' is the string date in format 'MM-DD-YYYY'
{
'home_name': Name of the home team, such as 'Purdue
Boilermakers' (`str`),
'home_abbr': Abbreviation for the home team, such as
'PURDUE' (`str`),
'away_name': Name of the away team, such as 'Indiana
Hoosiers' (`str`),
'away_abbr': Abbreviation for the away team, such as
'INDIANA' (`str`),
'boxscore': String representing the boxscore URI, such as
'2018-01-28-15-indiana' (`str`),
'non_di': Boolean value which evaluates to True when at
least one of the teams does not compete in NCAA
Division-I basketball (`bool`),
'top_25': Boolean value which evaluates to True when at
least one of the teams is ranked in the AP Top 25
polls (`bool`),
'winning_name': Full name of the winning team, such as
'Purdue Boilermakers' (`str`),
'winning_abbr': Abbreviation for the winning team, such as
'PURDUE' (`str`),
'losing_name': Full name of the losing team, such as
'Indiana Hoosiers' (`str`),
'losing_abbr': Abbreviation for the losing team, such as
'INDIANA' (`str`),
'home_score': Integer score for the home team (`int`),
'home_rank': Integer representing the home team's rank
(`int`),
'away_score': Integer score for the away team (`int`),
'away_rank': Integer representing the away team's rank
(`int`)
},
{ ... },
...
]
}
If no games were played on 'date', the list for ['date'] will be empty.
"""
return self._boxscores
def _create_url(self, date):
"""
Build the URL based on the passed datetime object.
In order to get the proper boxscore page, the URL needs to include the
requested month, day, and year.
Parameters
----------
date : datetime object
The date to search for any matches. The month, day, and year are
required for the search, but time is not factored into the search.
Returns
-------
string
Returns a ``string`` of the boxscore URL including the requested
date.
"""
return BOXSCORES_URL % (date.month, date.day, date.year)
def _get_requested_page(self, url):
"""
Get the requested page.
Download the requested page given the created URL and return a PyQuery
object.
Parameters
----------
url : string
The URL containing the boxscores to find.
Returns
-------
PyQuery object
A PyQuery object containing the HTML contents of the requested
page.
"""
return pq(url)
def _get_boxscore_uri(self, url):
"""
Find the boxscore URI.
Given the boxscore tag for a game, parse the embedded URI for the
boxscore.
Parameters
----------
url : PyQuery object
A PyQuery object containing the game's boxscore tag which has the
boxscore URI embedded within it.
Returns
-------
string
Returns a ``string`` containing the link to the game's boxscore
page.
"""
uri = re.sub(r'.*cfb/boxscores/', '', str(url))
uri = re.sub(r'\.html.*', '', uri).strip()
return uri
def _parse_abbreviation(self, abbr):
"""
Parse a team's abbreviation.
Given the team's HTML name tag, parse their abbreviation.
Parameters
----------
abbr : string
A string of a team's HTML name tag.
Returns
-------
string
Returns a ``string`` of the team's abbreviation.
"""
if 'cfb/schools' not in str(abbr):
return None
abbr = re.sub(r'.*/schools/', '', str(abbr))
abbr = re.sub(r'/.*', '', abbr)
return abbr
def _get_name(self, name):
"""
Find a team's name and abbreviation.
Given the team's HTML name tag, determine their name, abbreviation, and
whether or not they compete in Division-I.
Parameters
----------
name : PyQuery object
A PyQuery object of a team's HTML name tag in the boxscore.
Returns
-------
tuple
Returns a tuple containing the name, abbreviation, and whether or
not the team participates in Division-I. Tuple is in the following
order: Team Name, Team Abbreviation, boolean which evaluates to
True if the team does not participate in Division-I.
"""
team_name = name.text()
abbr = self._parse_abbreviation(name)
non_di = False
if not abbr:
abbr = team_name
non_di = True
return team_name, abbr, non_di
def _get_score(self, score_link):
"""
Find a team's final score.
Given an HTML string of a team's boxscore, extract the integer
representing the final score and return the number.
Parameters
----------
score_link : string
An HTML string representing a team's final score in the format
'<td class="right">NN</td>' where 'NN' is the team's score.
Returns
-------
int
Returns an int representing the team's final score in runs.
"""
score = score_link.replace('<td class="right">', '')
score = score.replace('</td>', '')
return int(score)
def _get_rank(self, team):
"""
Find the team's rank when applicable.
If a team is ranked, it will showup in a separate <span> tag with the
actual rank embedded between parentheses. When a team is ranked, the
integer value representing their ranking should be returned. For teams
that are not ranked, None should be returned.
Parameters
----------
team : PyQuery object
A PyQuery object of a team's HTML tag in the boxscore.
Returns
-------
int
Returns an integer representing the team's ranking when applicable,
or None if the team is not ranked.
"""
rank = None
rank_field = team('span[class="pollrank"]')
if len(rank_field) > 0:
rank = re.findall(r'\(\d+\)', str(rank_field))[0]
rank = int(rank.replace('(', '').replace(')', ''))
return rank
def _get_team_names(self, game):
"""
Find the names and abbreviations for both teams in a game.
Using the HTML contents in a boxscore, find the name and abbreviation
for both teams and determine wether or not this is a matchup between
two Division-I teams.
Parameters
----------
game : PyQuery object
A PyQuery object of a single boxscore containing information about
both teams.
Returns
-------
tuple
Returns a tuple containing the names and abbreviations of both
teams in the following order: Away Name, Away Abbreviation, Away
Score, Away Ranking, Home Name, Home Abbreviation, Home Score, Home
Ranking, a boolean which evaluates to True if either team does not
participate in Division-I athletics, and a boolean which evalutes
to True if either team is currently ranked.
"""
# Grab the first <td...> tag for each <tr> row in the boxscore,
# representing the name for each participating team.
links = [g('td:first') for g in game('tr').items()]
# The date will be included periodically in the boxscore which adds a
# third table row to the boxscore. In this case, the away team is the
# second link (index 1). Othertimes, the date is not included, and the
# first link is the away team.
if len(links) == 2:
away = links[0]
else:
away = links[1]
# The home team is the last (2nd or 3rd) link in the boxscore
home = links[-1]
non_di = False
scores = re.findall(r'<td class="right">\d+</td>', str(game))
away_score = None
home_score = None
# If the game hasn't started or hasn't been updated on sports-reference
# yet, no score will be shown and therefore can't be parsed.
if len(scores) == 2:
away_score = self._get_score(scores[0])
home_score = self._get_score(scores[1])
away_name, away_abbr, away_non_di = self._get_name(away('a'))
home_name, home_abbr, home_non_di = self._get_name(home('a'))
non_di = away_non_di or home_non_di
away_rank = self._get_rank(away)
home_rank = self._get_rank(home)
top_25 = bool(away_rank or home_rank)
return (away_name, away_abbr, away_score, away_rank, home_name,
home_abbr, home_score, home_rank, non_di, top_25)
def _get_team_results(self, away_name, away_abbr, away_score, home_name,
home_abbr, home_score):
"""
Determine the winner and loser of the game.
If the game has been completed and sports-reference has been updated
with the score, determine the winner and loser and return their
respective names and abbreviations.
Parameters
----------
away_name : string
The name of the away team, such as 'Indiana'.
away_abbr : string
The abbreviation of the away team, such as 'indiana'.
away_score : int
The number of points the away team scored, or None if the game
hasn't completed yet.
home_score : string
The name of the home team, such as 'Purdue'.
home_abbr : string
The abbreviation of the home team, such as 'purdue'.
home_score : int
The number of points the home team scored, or None if the game
hasn't completed yet.
Returns
-------
tuple, tuple
Returns two tuples, each containing the name followed by the
abbreviation of the winning and losing team, respectively. If the
game doesn't have a score associated with it yet, both tuples will
be None.
"""
if not away_score or not home_score:
return None, None
if away_score > home_score:
return (away_name, away_abbr), (home_name, home_abbr)
else:
return (home_name, home_abbr), (away_name, away_abbr)
def _extract_game_info(self, games):
"""
Parse game information from all boxscores.
Find the major game information for all boxscores listed on a
particular boxscores webpage and return the results in a list.
Parameters
----------
games : generator
A generator where each element points to a boxscore on the parsed
boxscores webpage.
Returns
-------
list
Returns a ``list`` of dictionaries where each dictionary contains
the name and abbreviations for both the home and away teams, a
boolean value indicating whether or not both teams compete in
Division-I, and a link to the boxscore.
"""
all_boxscores = []
for game in games:
names = self._get_team_names(game)
away_name, away_abbr, away_score, away_rank, home_name, \
home_abbr, home_score, home_rank, non_di, top_25 = names
boxscore_url = game('td[class="right gamelink"] a')
boxscore_uri = self._get_boxscore_uri(boxscore_url)
winning_name = None
winning_abbr = None
losing_name = None
losing_abbr = None
winner, loser = self._get_team_results(away_name, away_abbr,
away_score, home_name,
home_abbr, home_score)
if winner and loser:
winning_name, winning_abbr = winner
losing_name, losing_abbr = loser
game_info = {
'boxscore': boxscore_uri,
'away_name': away_name,
'away_abbr': away_abbr,
'away_score': away_score,
'away_rank': away_rank,
'home_name': home_name,
'home_abbr': home_abbr,
'home_score': home_score,
'home_rank': home_rank,
'non_di': non_di,
'top_25': top_25,
'winning_name': winning_name,
'winning_abbr': winning_abbr,
'losing_name': losing_name,
'losing_abbr': losing_abbr
}
all_boxscores.append(game_info)
return all_boxscores
def _find_games(self, date, end_date):
"""
Retrieve all major games played on a given day.
Builds a URL based on the requested date and downloads the HTML
contents before parsing any and all games played during that day. Any
games that are found are added to the boxscores dictionary with
high-level game information such as the home and away team names and a
link to the boxscore page.
Parameters
----------
date : datetime object
The date to search for any matches. The month, day, and year are
required for the search, but time is not factored into the search.
end_date : datetime object
Optionally specify an end date to iterate until. All boxscores
starting from the date specified in the 'date' parameter up to and
including the boxscores specified in the 'end_date' parameter will
be pulled. If left empty, or if 'end_date' is prior to 'date', only
the games from the day specified in the 'date' parameter will be
saved.
"""
# Set the end date to the start date if the end date is before the
# start date.
if not end_date or date > end_date:
end_date = date
date_step = date
while date_step <= end_date:
url = self._create_url(date_step)
page = self._get_requested_page(url)
games = page('table[class="teams"]').items()
boxscores = self._extract_game_info(games)
timestamp = '%s-%s-%s' % (date_step.month, date_step.day,
date_step.year)
self._boxscores[timestamp] = boxscores
date_step += timedelta(days=1)