def createTeamCodes(
years=["2008", "2009", "2010", "2011", "2012", "2013", "2014",
"2015"]):
initials = []
for year in years:
initials.extend(ScrapeWebsite.grab_team_initials(year))
allTeams = list(set(initials))
return allTeams
def createCSVPerSeason(years = ["2008","2009","2010","2011","2012","2013","2014","2015"]):
inPath = "C:\\NBA Data Science\\Scrapped\\"
newpath = "C:\\NBA Data Science\\Prepared\\"
if not os.path.exists(newpath): os.makedirs(newpath)
for year in years:
initials = ScrapeWebsite.grab_team_initials(year) #We'll be using functions inside our ScrapeWebsite
outputCSV = open(newpath + year + "_seasonDataSet.csv","w+")
outputCSV.write("date,win,tm0_init,tm0_MP,tm0_FG,tm0_FGA,tm0_FG%,tm0_3P,tm0_3PA,tm0_3P%,tm0_FT,tm0_FTA,tm0_FT%,tm0_ORB,tm0_DRB,tm0_TRB,tm0_AST,tm0_STL,tm0_BLK,tm0_TOV,tm0_PF,tm0_PTS,tm0_TS%,tm0_eFG%,tm0_3PAr,tm0_FTr,tm0_ORB%,tm0_DRB%,tm0_TRB%,tm0_AST%,tm0_STL%,tm0_BLK%,tm0_TOV%,tm0_USG%,tm0_ORtg,tm0_DRtg,tm1_init,tm1_MP,tm1_FG,tm1_FGA,tm1_FG%,tm1_3P,tm1_3PA,tm1_3P%,tm1_FT,tm1_FTA,tm1_FT%,tm1_ORB,tm1_DRB,tm1_TRB,tm1_AST,tm1_STL,tm1_BLK,tm1_TOV,tm1_PF,tm1_PTS,tm1_TS%,tm1_eFG%,tm1_3PAr,tm1_FTr,tm1_ORB%,tm1_DRB%,tm1_TRB%,tm1_AST%,tm1_STL%,tm1_BLK%,tm1_TOV%,tm1_USG%,tm1_ORtg,tm1_DRtg" + "\n")
for team in initials:
basicCSV = open(inPath + year + "\\" + team + "_basic.csv", "r").read()
advancedCSV = open(inPath + year + "\\" + team + "_advanced.csv","r").read()
outcomeCSV = open(inPath + year + "\\" + team + "_outcome.csv","r").read()
basicCSV = basicCSV.split('\n')
advancedCSV = advancedCSV.split('\n')
outcomeCSV = outcomeCSV.replace("G,Date,,,,,Opponent,,,Tm,Opp,W,L,Streak,Notes\n","") #Remove repeated headers from CSV
outcomeCSV = outcomeCSV.split('\n')
for i in range(0,len(basicCSV)):
if basicCSV[i] != "" and advancedCSV[i] != "" and outcomeCSV[i] != "":
#Create feature vectors
outcomeFeatures = outcomeCSV[i].split(',')
basicFeatures = basicCSV[i].split(',')
advFeatures = advancedCSV[i].split(',')
#Grab basic features for each team
team0_basic = basicFeatures[0:19]
team1_basic = basicFeatures[20:39]
#Grab advanced features for each team (Ignore the duplicated MP feature that's already inside our basic dataset)
team0_adv = advFeatures[1:15]
team1_adv = advFeatures[16:]
#Grab teamCode for opponent team by converting opponent's initial stored at the end of basicCSV
print team + " " + year
team1_initial = teamCodes.index(basicFeatures[40])
#Grab teamCode for current team
team0_initial = teamCodes.index(team)
#Create team0 and team1 entry
team0_entry = str(team0_initial) + "," + ','.join(team0_basic) + "," + ','.join(team0_adv)
team1_entry = str(team1_initial) + "," + ','.join(team1_basic) + "," + ','.join(team1_adv)
#Whether team0 won or lost
team0_win = outcomeFeatures[7].replace("W","1").replace("L","0") #We'll get other values if our data is corrupted
#Date of match between team0 and team1
date_match = outcomeFeatures[1]
#Add new data point to line in our seasonal data set
outputCSV.write(date_match + "," + team0_win + "," + team0_entry + "," + team1_entry + "\n")
#Close the file
outputCSV.close()
def createTeamCodes(years = ["2008","2009","2010","2011","2012","2013","2014","2015"]): initials = [] for year in years: initials.extend(ScrapeWebsite.grab_team_initials(year)) allTeams = list(set(initials)) return allTeams
def createCSVPerSeason(
years=["2008", "2009", "2010", "2011", "2012", "2013", "2014",
"2015"]):
inPath = "C:\\NBA Data Science\\Scrapped\\"
newpath = "C:\\NBA Data Science\\Prepared\\"
if not os.path.exists(newpath): os.makedirs(newpath)
for year in years:
initials = ScrapeWebsite.grab_team_initials(
year) #We'll be using functions inside our ScrapeWebsite
outputCSV = open(newpath + year + "_seasonDataSet.csv", "w+")
for team in initials:
basicCSV = open(inPath + year + "\\" + team + "_basic.csv",
"r").read()
advancedCSV = open(inPath + year + "\\" + team + "_advanced.csv",
"r").read()
outcomeCSV = open(inPath + year + "\\" + team + "_outcome.csv",
"r").read()
basicCSV = basicCSV.split('\n')
advancedCSV = advancedCSV.split('\n')
outcomeCSV = outcomeCSV.replace(
"G,Date,,,,,Opponent,,,Tm,Opp,W,L,Streak,Notes\n",
"") #Remove repeated headers from CSV
outcomeCSV = outcomeCSV.split('\n')
for i in range(0, len(basicCSV)):
if basicCSV[i] != "" and advancedCSV[i] != "" and outcomeCSV[
i] != "":
#Create feature vectors
outcomeFeatures = outcomeCSV[i].split(',')
basicFeatures = basicCSV[i].split(',')
advFeatures = advancedCSV[i].split(',')
#Grab basic features for each team
team0_basic = basicFeatures[0:19]
team1_basic = basicFeatures[20:39]
#Grab advanced features for each team (Ignore the duplicated MP feature that's already inside our basic dataset)
team0_adv = advFeatures[1:15]
team1_adv = advFeatures[16:]
#Grab teamCode for opponent team by converting opponent's initial stored at the end of basicCSV
print team + " " + year
team1_initial = teamCodes.index(basicFeatures[40])
#Grab teamCode for current team
team0_initial = teamCodes.index(team)
#Create team0 and team1 entry
team0_entry = str(team0_initial) + "," + ','.join(
team0_basic) + "," + ','.join(team0_adv)
team1_entry = str(team1_initial) + "," + ','.join(
team1_basic) + "," + ','.join(team1_adv)
#Whether team0 won or lost
team0_win = outcomeFeatures[7].replace("W", "1").replace(
"L",
"0") #We'll get other values if our data is corrupted
#Add new data point to line in our seasonal data set
outputCSV.write(team0_entry + "," + team1_entry + "," +
team0_win + "\n")
#Close the file
outputCSV.close()