def main(write_match):
c_data = csvreader.csv_reader_without_headers('../data/networkdata/match' +
str(write_match) + '_c.csv')
# c means clustering coefficient data.
# c[0] is player name, c[1] is the clustering coefficient value.
c = []
for item in c_data:
c.append([item[0], float(item[1])])
# First divide into 2 groups: Group A and others.
center, bel = kmeans(c, write_match, 2)
# Make center[1] group A.
if center[0] > center[1]:
center[0], center[1] = center[1], center[0]
for i, item in enumerate(bel):
if item == 1:
bel[i] = 0
else:
bel[i] = 1
# Copy other item to d[].
d = []
for i, item in enumerate(c):
if bel[i] == 0:
d.append(item)
# Divide others into 2 groups: Group B and Group C.
center2, bel2 = kmeans(d, write_match, 2)
# Make center2[0] Group C and center2[1] Group B.
if center2[0] > center2[1]:
center2[0], center2[1] = center2[1], center2[0]
# Return every group's center.
return [center2[0], center2[1], center[1]]
tnlocs = [nloc[i]]
dat = ps_data[key]
size = dat[1] * 20
alpha = dat[1] / 160 + 0.15
set_pts(graph,
tnodes,
tnlocs,
node_color=node_color,
size=size,
alpha=alpha)
length_factor = 1.05
width_factor = 0.68
full_events = csvreader.csv_reader_without_headers('../data/fullevents.csv')
passing_events = csvreader.csv_reader_without_headers(
'../data/passingevents.csv')
write_team = 'Huskies'
for write_match in range(1, 39):
location = {}
passing_data = {}
pair_passing_data = {}
substitution_list = []
for item in full_events:
if int(item[0]) is not write_match:
continue
if item[1].find(write_team) is -1:
def main(write_match, left_time, right_time):
full_events = csvreader.csv_reader_without_headers(
'../data/fullevents.csv')
passing_events = csvreader.csv_reader_without_headers(
'../data/passingevents.csv')
write_team = 'Huskies'
location = {}
name_list = []
# Count and average coordinates.
for item in full_events:
if int(item[0]) is not write_match:
continue
if item[1].find(write_team) is -1:
continue
tm = float(item[5])
if item[4].find('2') != -1:
tm += 3600
if left_time > tm or tm > right_time:
continue
name = (item[2].split('_'))[1]
if name not in name_list:
name_list.append(name)
if item[3] != '':
nm = (item[3].split('_'))[1]
if nm not in name_list:
name_list.append(nm)
if item[6].find('Pass') is not -1:
add_coordinates(location, name, float(item[8]), float(item[9]))
if item[3] != '' and item[3] != item[2]:
add_coordinates(location, (item[3].split('_'))[1],
float(item[10]), float(item[11]))
elif item[6].find('Save') is not -1 or item[6].find(
'Goalkeeper') is not -1:
add_coordinates(location, name, 0, 50)
elif item[6].find('Duel') is not -1 or item[6].find('Others') is not -1 or \
item[6].find('Foul') is not -1 or item[6].find('Offside') is not -1 or item[6].find('Shot') is not -1:
add_coordinates(location, name, float(item[8]), float(item[9]))
# Compute the player's average location.
for key in location:
loc = location[key]
assert loc[2] is not 0
location[key] = [loc[0] / loc[2], loc[1] / loc[2]]
# Compute the average coordinate distance between players.
dC = {}
for name1 in name_list:
for name2 in name_list:
if name1 == name2:
continue
key = name1 + '.' + name2
if location.get(name1) is None:
if name1.find('G') != -1:
location[name1] = [0, 50]
else:
continue
if location.get(name2) is None:
if name2.find('G') != -1:
location[name2] = [0, 50]
else:
continue
dC[key] = euclid_distance_loc(location[name1], location[name2])
# Count the times of each pass and passing times distance
head_pass = {}
hand_pass = {}
high_pass = {}
simple_pass = {}
launch = {}
cross = {}
smart_pass = {}
dT = {}
for name1 in name_list:
for name2 in name_list:
if name1 == name2:
continue
key = name1 + '.' + name2
dT[key] = 0
for item in passing_events:
if int(item[0]) is not write_match:
continue
if item[1].find(write_team) is -1:
continue
tm = float(item[5])
if item[4].find('2') != -1:
tm += 3600
if left_time > tm or tm > right_time:
continue
name1 = (item[2].split('_'))[1]
name2 = (item[3].split('_'))[1]
if name1 == name2:
continue
key = name1 + '.' + name2
dT[key] += 1
dis = euclid_distance(float(item[7]), float(item[8]), float(item[9]),
float(item[10]))
if item[6].find('Head') != -1: # Head pass
add_counts(head_pass, key, dis)
elif item[6].find('Hand') != -1: # Hand pass
add_counts(hand_pass, key, dis)
elif item[6].find('High') != -1: # High pass
add_counts(high_pass, key, dis)
elif item[6].find('Simple') != -1: # Simple pass
add_counts(simple_pass, key, dis)
elif item[6].find('Launch') != -1: # Launch
add_counts(launch, key, dis)
elif item[6].find('Cross') != -1: # Cross
add_counts(cross, key, dis)
elif item[6].find('Smart') != -1: # Smart pass
add_counts(smart_pass, key, dis)
# Compute the average passing distance.
dP = {}
kdP = {
'Head pass': 0.8,
'Hand pass': 0.6,
'High pass': 1.3,
'Simple pass': 1.0,
'Launch': 0.7,
'Cross': 1.2,
'Smart pass': 1.1
}
kdP_total = kdP['Head pass'] + kdP['Hand pass'] + kdP['High pass'] + kdP['Simple pass'] + \
kdP['Cross'] + kdP['Smart pass'] + kdP['Launch']
for name1 in name_list:
for name2 in name_list:
if name1 == name2:
continue
key = name1 + '.' + name2
score = get_counts(head_pass, key) * kdP['Head pass'] \
+ get_counts(hand_pass, key) * kdP['Hand pass'] \
+ get_counts(high_pass, key) * kdP['High pass'] \
+ get_counts(simple_pass, key) * kdP['Simple pass'] \
+ get_counts(launch, key) * kdP['Launch'] \
+ get_counts(cross, key) * kdP['Cross'] \
+ get_counts(smart_pass, key) * kdP['Smart pass']
score = score / kdP_total
dP[key] = score
# Compute dT and ready for standardize
maxdC = 0
maxdP = 0
for name1 in name_list:
for name2 in name_list:
if name1 == name2:
continue
key = name1 + '.' + name2
if dT[key] == 0:
continue
dT[key] = 1 / dT[key]
maxdC = max(maxdC, dC[key])
maxdP = max(maxdP, dP[key])
# Standardize & Compute w
w = {}
for name1 in name_list:
for name2 in name_list:
if name1 == name2:
continue
key = name1 + '.' + name2
if dT[key] == 0:
continue
dC[key] = dC[key] / maxdC
dP[key] = dP[key] / maxdP
w[key] = 1 / (0.7 * dT[key] + 0.2 * dC[key] + 0.1 * dP[key])
w[key] = math.exp(w[key])
a = {}
k = {}
for name1 in name_list:
count = 0
for name2 in name_list:
key = name1 + '.' + name2
if w.get(key) is None:
a[key] = 0
else:
a[key] = 1
count += 1
k[name1] = count
c_count = 0
c_num = 0
c = {}
for name in name_list:
sum = 0
for name2 in name_list:
for name3 in name_list:
key1 = name2 + '.' + name
key2 = name + '.' + name3
key3 = name2 + '.' + name3
if a[key1] == 1 and a[key2] == 1 and a[key3] == 1:
sum = sum + (w[key1] + w[key2]) / 2
if k[name] != 0 and k[name] != 1:
c[name] = sum / k[name] / (k[name] - 1)
c_count += c[name]
c_num += 1
# print(name, c[name])
else:
c[name] = 0
return c
def main(write_match):
full_events = csvreader.csv_reader_without_headers(
'../data/fullevents.csv')
passing_events = csvreader.csv_reader_without_headers(
'../data/passingevents.csv')
write_team = 'Huskies'
location = {}
substitution_list = []
name_list = []
dT = {}
for item in full_events:
if int(item[0]) is not write_match:
continue
if item[1].find(write_team) is -1:
continue
name = (item[2].split('_'))[1]
if name not in name_list:
name_list.append(name)
for name1 in name_list:
for name2 in name_list:
if name1 == name2:
continue
key = name1 + '.' + name2
dT[key] = 0
for item in passing_events:
if int(item[0]) is not write_match:
continue
if item[1].find(write_team) is -1:
continue
name1 = (item[2].split('_'))[1]
name2 = (item[3].split('_'))[1]
if name1 == name2:
continue
key = name1 + '.' + name2
dT[key] += 1
graph = pr.Graph()
for name in name_list:
sum = 0
for name2 in name_list:
if name == name2:
continue
key = name + '.' + name2
sum += dT[key]
for name2 in name_list:
if name == name2:
continue
key = name + '.' + name2
c = dT[key] / sum
graph.add_link(name, name2, c)
pr_map = graph.get_PR()
with open('../data/networkdata/match' + str(write_match) + '_pr.csv',
'w',
newline='') as csv_file:
csv_writer = csv.writer(csv_file, dialect='excel')
for key in pr_map:
csv_writer.writerow([key, pr_map[key]])
def main(write_match):
full_events = csvreader.csv_reader_without_headers(
'../data/fullevents.csv')
passing_events = csvreader.csv_reader_without_headers(
'../data/passingevents.csv')
# Only consider our team: Huskies.
write_team = 'Huskies'
location = {}
substitution_list = []
name_list = []
# Count substitutions and average coordinates of players.
for item in full_events:
if int(item[0]) is not write_match:
continue
if item[1].find(write_team) is -1:
continue
name = (item[2].split('_'))[1]
if name not in name_list:
name_list.append(name)
if item[6].find('Pass') is not -1:
add_coordinates(location, name, float(item[8]), float(item[9]))
if item[3] is not '' and item[3] != item[2]:
add_coordinates(location, (item[3].split('_'))[1],
float(item[10]), float(item[11]))
elif item[6].find('Save') is not -1 or item[6].find(
'Goalkeeper') is not -1:
add_coordinates(location, name, 0, 50)
elif item[6].find('Substitution') is not -1:
substitution_list.append((item[3].split('_'))[1])
elif item[6].find('Duel') is not -1 or item[6].find('Others') is not -1 or \
item[6].find('Foul') is not -1 or item[6].find('Offside') is not -1 or \
item[6].find('Shot') is not -1:
add_coordinates(location, name, float(item[8]), float(item[9]))
# Compute the player's average location.
for key in location:
loc = location[key]
assert loc[2] is not 0
location[key] = [loc[0] / loc[2], loc[1] / loc[2]]
# Compute the average coordinate distance between players.
dC = {}
for name1 in name_list:
for name2 in name_list:
if name1 == name2:
continue
key = name1 + '.' + name2
dC[key] = euclid_distance_loc(location[name1], location[name2])
# Count the times of each pass and passing times distance
head_pass = {}
hand_pass = {}
high_pass = {}
simple_pass = {}
launch = {}
cross = {}
smart_pass = {}
dT = {}
for name1 in name_list:
for name2 in name_list:
if name1 == name2:
continue
key = name1 + '.' + name2
dT[key] = 0
for item in passing_events:
if int(item[0]) is not write_match:
continue
if item[1].find(write_team) is -1:
continue
name1 = (item[2].split('_'))[1]
name2 = (item[3].split('_'))[1]
if name1 == name2:
continue
key = name1 + '.' + name2
dT[key] += 1
# Use euclid distance.
dis = euclid_distance(float(item[7]), float(item[8]), \
float(item[9]), float(item[10]))
if item[6].find('Head') != -1: # Head pass
add_counts(head_pass, key, dis)
elif item[6].find('Hand') != -1: # Hand pass
add_counts(hand_pass, key, dis)
elif item[6].find('High') != -1: # High pass
add_counts(high_pass, key, dis)
elif item[6].find('Simple') != -1: # Simple pass
add_counts(simple_pass, key, dis)
elif item[6].find('Launch') != -1: # Launch
add_counts(launch, key, dis)
elif item[6].find('Cross') != -1: # Cross
add_counts(cross, key, dis)
elif item[6].find('Smart') != -1: # Smart pass
add_counts(smart_pass, key, dis)
# Compute the average passing distance.
dP = {}
kdP = {
'Head pass': 0.8,
'Hand pass': 0.6,
'High pass': 1.3,
'Simple pass': 1.0,
'Launch': 0.7,
'Cross': 1.2,
'Smart pass': 1.1
}
kdP_total = kdP['Head pass'] + kdP['Hand pass'] + kdP['High pass'] + \
kdP['Simple pass'] + kdP['Cross'] + kdP['Smart pass'] + kdP['Launch']
for name1 in name_list:
for name2 in name_list:
if name1 == name2:
continue
key = name1 + '.' + name2
score = get_counts(head_pass, key) * kdP['Head pass'] \
+ get_counts(hand_pass, key) * kdP['Hand pass'] \
+ get_counts(high_pass, key) * kdP['High pass'] \
+ get_counts(simple_pass, key) * kdP['Simple pass'] \
+ get_counts(launch, key) * kdP['Launch'] \
+ get_counts(cross, key) * kdP['Cross'] \
+ get_counts(smart_pass, key) * kdP['Smart pass']
score = score / kdP_total
dP[key] = score
# Compute dT and ready for standardize
maxdC = 0
maxdP = 0
for name1 in name_list:
for name2 in name_list:
if name1 == name2:
continue
key = name1 + '.' + name2
if dT[key] == 0:
continue
dT[key] = 1 / dT[key]
maxdC = max(maxdC, dC[key])
maxdP = max(maxdP, dP[key])
# Standardize & Compute weight of each edge.
w = {}
for name1 in name_list:
for name2 in name_list:
if name1 == name2:
continue
if name1 in substitution_list or name2 in substitution_list:
continue
key = name1 + '.' + name2
if dT[key] == 0:
continue
# Standardize.
dC[key] = dC[key] / maxdC
dP[key] = dP[key] / maxdP
# Calculate weight of each edge.
w[key] = 1 / (0.7 * dT[key] + 0.2 * dC[key] + 0.1 * dP[key])
w[key] = math.exp(w[key])
# Write weight result to an file.
headers = ['From', 'To', 'w']
file_name = '../data/networkdata/match' + str(write_match) + '_w.csv'
with open(file_name, 'w', newline='') as csv_file:
csv_writer = csv.writer(csv_file, dialect='excel')
csv_writer.writerow(headers)
for key in w:
[name1, name2] = key.split('.')
csv_writer.writerow([name1, name2, w[key]])
# Calculate Clustering Coefficient of every player.
a = {}
k = {}
for name1 in name_list:
count = 0
for name2 in name_list:
key = name1 + '.' + name2
if w.get(key) is None:
a[key] = 0
else:
a[key] = 1
count += 1
k[name1] = count
c_count = 0
c_num = 0
c = {}
for name in name_list:
sum = 0
for name2 in name_list:
for name3 in name_list:
key1 = name2 + '.' + name
key2 = name + '.' + name3
key3 = name2 + '.' + name3
if a[key1] == 1 and a[key2] == 1 and a[key3] == 1:
sum = sum + (w[key1] + w[key2]) / 2
if k[name] != 0 and k[name] != 1:
c[name] = sum / k[name] / (k[name] - 1)
c_count += c[name]
c_num += 1
else:
c[name] = 0
# Write result to an file.
headers = ['Name', 'Clustering Coefficient']
file_name = '../data/networkdata/match' + str(write_match) + '_c.csv'
with open(file_name, 'w', newline='') as csv_file:
csv_writer = csv.writer(csv_file, dialect='excel')
csv_writer.writerow(headers)
for key in c:
if c[key] is not 0:
csv_writer.writerow([key, c[key]])