def saveToFile(entropyLeagueDict, leagues):
featureVector = []
idNameMap = mapLeagueIdtoName(leagues)
data = []
for league in leagues:
leagueRow = []
entropyMap = entropyLeagueDict.get(league[0])
if entropyMap is None: continue
featureVector.append(idNameMap.get(league[0]))
leagueRow.append(entropyMap.get('2008/2009'))
leagueRow.append(entropyMap.get('2009/2010'))
leagueRow.append(entropyMap.get('2010/2011'))
leagueRow.append(entropyMap.get('2011/2012'))
leagueRow.append(entropyMap.get('2012/2013'))
leagueRow.append(entropyMap.get('2013/2014'))
leagueRow.append(entropyMap.get('2014/2015'))
leagueRow.append(entropyMap.get('2015/2016'))
data.append(leagueRow)
data = np.asarray(data)
utils.createFile(data.T, "seasonEntropy.csv", featureVector)
def DimRed(temp,algo,namesRating):
if algo == 'PCA':
pca = PCA(n_components=2,random_state=0).fit_transform(temp)
else:
pca = TSNE(n_components=2,random_state=0).fit_transform(temp)
comp1 = pca[:,0]
comp2 = pca[:,1]
data = []
for a,b,c in zip(comp1,comp2,namesRating):
data.append([a,b,c[0],c[1]])
data = [row for row in data if row[3]>80]
data.sort(key = lambda x: x[3],reverse=True)
#data =[[a,b] for a,b in zip(comp1,comp2)]
featureVector = ["pc1","pc2","name","rating"]
if algo == 'PCA':
utils.createFile(data, "components.csv",featureVector)
else:
utils.createFile(data, "TSNEcomponents.csv",featureVector)
def getLeagueTeamEntropyDict(matchEntropies, filteredMatches, leagues, teams):
print(teams[0])
teamMap = mapTeamIdtoName(teams)
leagueMap = mapLeagueIdtoName(leagues)
leagueTeamEntropyDict = dict()
for match in filteredMatches:
leagueTeamEntropyDict[match[2]] = dict()
for match in filteredMatches:
league = match[2]
homeTeam = match[7]
awayTeam = match[8]
season = match[3]
if leagueTeamEntropyDict[league].get(homeTeam) is None:
leagueTeamEntropyDict[league][homeTeam] = dict()
if leagueTeamEntropyDict[league].get(homeTeam).get(season) is None:
leagueTeamEntropyDict[league][homeTeam][season] = []
if leagueTeamEntropyDict[league].get(awayTeam) is None:
leagueTeamEntropyDict[league][awayTeam] = dict()
if leagueTeamEntropyDict[league].get(awayTeam).get(season) is None:
leagueTeamEntropyDict[league][awayTeam][season] = []
for entropy, match in zip(matchEntropies, filteredMatches):
leagueTeamEntropyDict[match[2]][match[7]][match[3]].append(entropy)
leagueTeamEntropyDict[match[2]][match[8]][match[3]].append(entropy)
finalData = []
for league in leagueTeamEntropyDict.keys():
teams = leagueTeamEntropyDict.get(league).keys()
for team in teams:
seasons = leagueTeamEntropyDict.get(league).get(team).keys()
for season in seasons:
meanEntropy = np.mean(
leagueTeamEntropyDict.get(league).get(team).get(season))
teamName = teamMap.get(team).split(",")[0]
finalData.append(
[teamName, season, meanEntropy,
leagueMap.get(league)])
featureVector = ["Team", "Season", "MeanEntropy", "League"]
utils.createFile(finalData, "teamLeagueSeasonEntropy.csv", featureVector)
def getEvolutionData(players):
#print (players[0])
retVal = dict()
sortedByRating = sorted(players, key = lambda x : x[1],reverse=True)[:10]
for player in sortedByRating:
playerMap = dict()
maptoAttach = dict()
playerName = player[-2]
x = player[0]
temp = []
for row in Database().execute('select date,overall_rating from Player_Attributes where player_api_id is '+str(x)):
temp.append(row)
for row in temp:
if playerMap.get(row[0].split('-')[0]) is None:
playerMap[row[0].split('-')[0]] = []
playerMap[row[0].split('-')[0]].append(row[1])
#print(playerMap)
retVal[playerName] = dict()
for k,v in playerMap.items():
retVal[playerName][k]=np.mean(v)
#print(retVal)
data = []
featureVector = []
for player in retVal.keys():
row = []
featureVector.append(player)
row.append(retVal.get(player).get('2007'))
row.append(retVal.get(player).get('2008'))
row.append(retVal.get(player).get('2009'))
row.append(retVal.get(player).get('2010'))
row.append(retVal.get(player).get('2011'))
row.append(retVal.get(player).get('2012'))
row.append(retVal.get(player).get('2013'))
row.append(retVal.get(player).get('2014'))
row.append(retVal.get(player).get('2015'))
data.append(row)
data = np.asarray(data)
utils.createFile(data.T,"topPlayerRating.csv",featureVector)
matches = utils.getAllDatafromTable('match')
leagueScoreMap = dict()
for league in leagues:
leagueScoreMap[league[0]] = []
for match in matches:
home_team_goals = int(match[9])
away_team_goals = int(match[10])
if home_team_goals == away_team_goals:
home_team_score = 1
away_team_score = 1
elif home_team_goals > away_team_goals:
home_team_score = 3
away_team_score = 0
else:
away_team_score = 3
home_team_score = 0
leagueScoreMap[match[2]].append([home_team_score, away_team_score])
avgLeagueScore = dict()
for league in leagueScoreMap.keys():
scores = leagueScoreMap.get(league)
avg_home_score = np.mean([score[0] for score in scores])
avg_away_score = np.mean([score[1] for score in scores])
avgLeagueScore[league] = [avg_home_score, avg_away_score]
data = []
for league in avgLeagueScore.keys():
avg_home_score = avgLeagueScore.get(league)[0]
avg_away_score = avgLeagueScore.get(league)[1]
countryName = countryNameMap.get(league)
data.append([countryName, avg_home_score, avg_away_score])
utils.createFile(data, "homeAdvantage.csv",
["country", "avg_home_score", "avg_away_score"])
from lib import utils
leagues = utils.getAllDatafromTable('league')
matches = utils.getAllDatafromTable('match')
countries = utils.getAllDatafromTable('country')
leagueIdMap = dict()
leagueTeamMap = dict()
leagueCountryMap = dict()
for league in leagues:
leagueIdMap[league[0]] = league[2]
for league in leagues:
leagueCountryMap[league[1]] = league[0]
#matches = utils.removeNan(matches)
for match in matches:
if leagueTeamMap.get(match[2]) is None: leagueTeamMap[match[2]] = set()
#print(match[2])
leagueTeamMap[match[2]].add(match[7])
leagueTeamMap[match[2]].add(match[8])
#print(leagueTeamMap.keys())
data = []
for country in countries:
leagueId = leagueCountryMap.get(country[0])
#print(leagueId)
leagueName = leagueIdMap.get(leagueId)
countryName = country[1]
numTeams = len(leagueTeamMap.get(leagueId))
data.append([countryName, leagueName, numTeams])
utils.createFile(data, "euroMap.csv", ["country", "league", "# Teams"])