def build_csv(overwrite):
if not os.path.exists('../data/p2p_assists/'):
os.makedirs('../data/p2p_assists')
print('MAKING DIRECTORY : ' + './data/p2p_assists')
file_path = '../data/p2p_assists/passing.csv'
if (not os.path.isfile(file_path)) or overwrite:
passing_df = pd.DataFrame(
columns=['Year', 'Receiver_Id', 'Receiver_Name', 'Passer_Id', 'Passer_Name', 'Frequency', 'Pass',
'Assists'])
for year in range(2013, 2016):
base_df = get_general_stats(PlayerOrTeam.P, MeasureTypes.BASE, PerModes.TOTAL, get_year_string(year),
SeasonTypes.REG)
for index, receiver in base_df.iterrows():
print(receiver.PLAYER_NAME)
player_passing_df = get_player_passing_dashboard(receiver.PLAYER_ID, get_year_string(year))
if player_passing_df is not None:
for index2, passer in player_passing_df.iterrows():
row = [year, passer.PLAYER_ID, passer.PLAYER_NAME_LAST_FIRST, passer.PASS_TEAMMATE_PLAYER_ID,
passer.PASS_FROM, passer.FREQUENCY, passer.PASS, passer.AST]
passing_df = passing_df.append(pd.Series(row, index=passing_df.columns), ignore_index=True)
passing_df = passing_df.sort_values(by='Assists', ascending=False)
passing_df.to_csv(file_path)
return passing_df
else:
return pd.read_csv(file_path)
def make_matplot_kde_shot_chart(df, player_name, player_id, year, season_type, chart_type):
player_pic = get_player_picture(player_id, player_name)
# create our jointplot
# get our colormap for the main kde plot
# Note we can extract a color from cmap to use for
# the plots that lie on the side and top axes
cmap = plt.cm.viridis
# n_levels sets the number of contour lines for the main kde plot
joint_shot_chart = sns.jointplot(df.LOC_X, df.LOC_Y, stat_func=None,
kind='kde', space=0, color=cmap(0.1),
cmap=cmap, n_levels=50, extent=[-250, 250, 422.5, -47.5])
joint_shot_chart.fig.set_size_inches(12, 11)
# A joint plot has 3 Axes, the first one called ax_joint
# is the one we want to draw our court onto and adjust some other settings
ax = joint_shot_chart.ax_joint
draw_court(ax)
# Adjust the axis limits and orientation of the plot in order
# to plot half court, with the hoop by the top of the plot
ax.set_xlim(-250, 250)
ax.set_ylim(422.5, -47.5)
# Get rid of axis labels and tick marks
ax.set_xlabel('')
ax.set_ylabel('')
ax.tick_params(labelbottom='off', labelleft='off')
# Add a title
title = player_name + ' ' + dg.get_year_string(year) + ' ' + season_type
ax.set_title(title,
y=1.2, fontsize=18)
# Add Data Scource and Author
ax.text(-250, 445, 'Data Source: stats.nba.com',
fontsize=12)
# Add Harden's image to the top right
# First create our OffSetImage by passing in our image
# and set the zoom level to make the image small enough
# to fit on our plot
img = OffsetImage(player_pic, zoom=0.6)
# Pass in a tuple of x,y coordinates to set_offset
# to place the plot where you want, I just played around
# with the values until I found a spot where I wanted
# the image to be
img.set_offset((987, 907))
# add the image
ax.add_artist(img)
file_path = '../charts/' + chart_type + '/kde/' + player_name + '_' + str(
year) + '_' + season_type + '.png'
dg.create_directories_and_check_for_file(file_path)
plt.savefig(file_path)
plt.close()
def make_histogram(df, player_name, year, season_type, chart_type):
plt.figure()
title = player_name + ' ' + dg.get_year_string(year) + ' ' + season_type
plt.title(title)
plt.xlabel('Distance (Feet)')
df['SHOT_DISTANCE'].plot.hist(alpha=0.5)
file_path = '../charts/' + chart_type + '/hist/' + player_name + '_' + str(year) + '_' + season_type + '.png'
dg.create_directories_and_check_for_file(file_path)
plt.savefig(file_path)
plt.close()
def get_season_game_ids(season_year):
url = (
"http://stats.nba.com/stats/leaguegamelog?"
"Counter=1000&"
"DateFrom=&"
"DateTo=&"
"Direction=DESC&"
"LeagueID=00&"
"PlayerOrTeam=T&"
"Season={seasonYear}&"
"SeasonType=Regular+Season&"
"Sorter=PTS"
)
url = url.format(seasonYear=data.get_year_string(season_year))
func_df = data.json_to_pandas(url, 0)
return func_df["GAME_ID"].unique()
def make_matplot_hexbin_shot_chart(df, player_name, player_id, year, season_type, chart_type):
player_pic = get_player_picture(player_id, player_name)
# create our jointplot
cmap = plt.cm.coolwarm
plt.axis([-250, 250, 422.5, -47.5])
joint_shot_chart = sns.jointplot(df.LOC_X, df.LOC_Y, stat_func=None, reduce_C_function=np.sum,
kind='hex', space=0, color=cmap(.2), cmap=cmap, extent=[-250, 250, 422.5, -47.5],
gridsize=30)
joint_shot_chart.fig.set_size_inches(12, 11)
# A joint plot has 3 Axes, the first one called ax_joint
# is the one we want to draw our court onto
ax = joint_shot_chart.ax_joint
draw_court(ax)
# Adjust the axis limits and orientation of the plot in order
# to plot half court, with the hoop by the top of the plot
ax.set_xlim(-250, 250)
ax.set_ylim(422.5, -47.5)
# Get rid of axis labels and tick marks
ax.set_xlabel('')
ax.set_ylabel('')
ax.tick_params(labelbottom='off', labelleft='off')
# Add a title
ax.set_title(player_name + ' ' + dg.get_year_string(year) + ' ' + season_type,
y=1.2, fontsize=18)
img = OffsetImage(player_pic, zoom=0.6)
img.set_offset((987, 907))
ax.add_artist(img)
file_path = '../charts/' + chart_type + '/hexbin/' + player_name + '_' + str(
year) + '_' + season_type + '.png'
dg.create_directories_and_check_for_file(file_path)
plt.savefig(file_path)
plt.close()
def make_matplot_scatter_shot_chart(df, player_name, player_id, year, season_type, chart_type):
player_pic = get_player_picture(player_id, player_name)
# create our jointplot
joint_shot_chart = sns.jointplot(df.LOC_X, df.LOC_Y, stat_func=None, kind='scatter', space=0, alpha=0.5)
joint_shot_chart.fig.set_size_inches(12, 11)
# A joint plot has 3 Axes, the first one called ax_joint
# is the one we want to draw our court onto and adjust some other settings
ax = joint_shot_chart.ax_joint
draw_court(ax)
# Adjust the axis limits and orientation of the plot in order
# to plot half court, with the hoop by the top of the plot
ax.set_xlim(-250, 250)
ax.set_ylim(422.5, -47.5)
# Get rid of axis labels and tick marks
ax.set_xlabel('')
ax.set_ylabel('')
ax.tick_params(labelbottom='off', labelleft='off')
# Add a title
ax.set_title(player_name + ' ' + dg.get_year_string(year) + ' ' + season_type,
y=1.2, fontsize=18)
img = OffsetImage(player_pic, zoom=.7)
img.set_offset((987, 907))
ax.add_artist(img)
file_path = '../charts/' + chart_type + '/scatter/' + player_name + '_' + str(
year) + '_' + season_type + '.png'
dg.create_directories_and_check_for_file(file_path)
plt.savefig(file_path)
plt.close()
def merge_pbp_and_shot_data(year):
file_path = "../data/merged_shot_pbp/" + str(year) + ".csv"
if not data.create_directories_and_check_for_file(file_path):
game_ids = get_season_game_ids(year)
full_year_merged_df = pd.DataFrame()
shot_df = data.get_shot_data(overwrite=False, season_year=year)
shot_df = shot_df[
[
"ACTION_TYPE",
"EVENT_TYPE",
"GAME_EVENT_ID",
"GAME_ID",
"LOC_X",
"LOC_Y",
"SHOT_DISTANCE",
"SHOT_MADE_FLAG",
"SHOT_TYPE",
"SHOT_ZONE_AREA",
"SHOT_ZONE_BASIC",
"SHOT_ZONE_RANGE",
]
]
for i, game_id in enumerate(game_ids):
print(i, game_id)
pbp_df = data.get_pbp_data(game_id, data.get_year_string(year), data.SeasonTypes.REG, overwrite=False)
try:
pbp_df = pbp_df[
[
"GAME_ID",
"EVENTNUM",
"PERIOD",
"PCTIMESTRING",
"HOMEDESCRIPTION",
"NEUTRALDESCRIPTION",
"VISITORDESCRIPTION",
"PLAYER1_ID",
"PLAYER1_NAME",
"PLAYER1_TEAM_ID",
"PLAYER2_ID",
"PLAYER2_NAME",
"PLAYER2_TEAM_ID",
"PLAYER3_ID",
"PLAYER3_NAME",
"PLAYER3_TEAM_ID",
]
]
merge_df = pd.merge(
pbp_df, shot_df, left_on=["GAME_ID", "EVENTNUM"], right_on=["GAME_ID", "GAME_EVENT_ID"], how="inner"
)
except KeyError:
print("KEY ERROR")
full_year_merged_df = full_year_merged_df.append(merge_df)
full_year_merged_df.to_csv("../data/merged_shot_pbp/" + str(year) + ".csv")
return full_year_merged_df
else:
return pd.read_csv(file_path)
import MySQLdb
import data_getters as data
import pandas as pd
db = MySQLdb.connect(host="localhost", user="******", passwd="1qw2SQL3er4", db="nbadb")
cur = db.cursor()
df = pd.DataFrame()
for year in range(1996, 2016):
base_df = data.get_general_stats(
data.PlayerOrTeam.P, data.MeasureTypes.BASE, data.PerModes.TOTAL, year, data.SeasonTypes.REG
)
base_df["YEAR"] = data.get_year_string(year)
base_df = base_df[
[
"PLAYER_ID",
"TEAM_ID",
"YEAR",
"AGE",
"PLAYER_NAME",
"TEAM_ABBREVIATION",
"GP",
"MIN",
"AST",
"BLK",
"DREB",
"OREB",
"FGA",
"FGM",
"FG3A",
