train_prediction.py

import sklearn
import numpy as np
import csv
from collections import defaultdict

from sklearn.linear_model import LinearRegression


SHOOTER_IDX = 0
CONTESTED_IDX = 1
NDD_IDX = 2
RELEASE_ANGLE_IDX = 3
CAS_IDX = 4

X_Train_dict = defaultdict(list)

GAME_IDS = [
    "M_8279c1d6-4e02-11ea-ab59-024282923f19",
    "M_9672b049-537c-11ea-ab59-024282923f19",
    "M_17472065-4ad8-11ea-9084-0242bdc61da9",
    "M_a8114fce-5374-11ea-ab59-024282923f19",
    "M_aa5ba87c-4d2a-11ea-b7ea-0242bdc61da9",
    "M_ea97103a-537c-11ea-ab59-024282923f19"
]

def read_shooter_csv(file_name):
    csv_file = {}
    with open(file_name) as csvfile:
        reader = csv.reader(csvfile)
        first_row = True
        for row in reader:
            if first_row:
                first_row = False
                continue
            csv_file[int(row[0])] = [float(row[1]), float(row[-1])]
    return csv_file

def read_defender_csv(file_name):
    csv_file = {}
    with open(file_name) as csvfile:
        reader = csv.reader(csvfile)
        first_row = True
        for row in reader:
            if first_row:
                first_row = False
                continue
            csv_file[int(row[0])] = float(row[1])
    return csv_file

def mean(samples):
    if not samples:
        return 0
    return sum(samples) / float(len(samples))


def read_csv(file_name):
    csv_file = []
    with open(file_name) as csvfile:
        reader = csv.reader(csvfile)
        for row in reader:
            csv_file.append(row)
    return csv_file

shooter_table = read_shooter_csv("shooter_table_mens.csv")
defender_table = read_defender_csv("defender_table_mens.csv")

for game_id in GAME_IDS:
    # game_id = "M_17472065-4ad8-11ea-9084-0242bdc61da9"
    shot_table_file = read_csv("shot_table_" + game_id + ".csv")
    for shot in shot_table_file[1:]:
        shooter_id = int(shot[SHOOTER_IDX])
        if shot[NDD_IDX] != 'NA':
            ndd_id = int(float(shot[NDD_IDX]))
        contested = int(shot[CONTESTED_IDX])
        angle = float(shot[RELEASE_ANGLE_IDX])
        if angle < 10:
            continue
        cas = (shot[CAS_IDX])

        if shooter_id not in shooter_table:
            print (shooter_id, "not in shooter table")
            continue

        if contested:
            shooter_avg_angle = shooter_table[shooter_id][0]
            defender_score = defender_table[ndd_id]
            shooter_score = shooter_table[shooter_id][1]
            # print ("shooter_score", shooter_score)
            #
            # print ("defender_score", defender_score)
            if shooter_score != -1:
                X_Train_dict[(shooter_score, defender_score)].append(abs(shooter_avg_angle - angle))

X = []
y = []

i = 0
for example in X_Train_dict:

    samples = X_Train_dict[example]
    sample_mean = mean(samples)

    X.append([1, example[0], example[1]])
    y.append(sample_mean)
    i+=1

X = np.array(X)
y = np.array(y)


reg = LinearRegression().fit(X, y)

print(reg.predict(np.array([[1.0, 0.15, 9.2]])))