search

anthonyshook / dfsOptimizer

R-Package for Optimizing DFS Slates
https://anthonyshook.com/dfsOptimizer/
Other
3 stars 1 forks source link

lock_players_by_id() and set_player_min_exp() not working for draftkings nba classic optimizer #13

Closed NahomKun closed 2 years ago

NahomKun commented 2 years ago

the optimizer doesn't lock the players to lock nor does it construct lineups with players with the minimum amount of player exposure df.csv code: `dkopt <- create_optimizer(site = 'DraftKings', sport = 'BASKETBALL', contest_type = 'CLASSIC')

dkopt <- add_players_from_df(object = dkopt, df = df)

min <- get_player_id(dkopt,'Cam Reddish') #SG/SF min2 <-get_player_id(dkopt,'Willy Hernangomez') #PF/C min3 <-get_player_id(dkopt,'Delon Wright') #PG min4 <-get_player_id(dkopt,'Saddiq Bey') #SF/SG min5 <-get_player_id(dkopt,'LeBron James') #SF/PF

dkopt<-dkopt%>% lock_players_by_id(min)%>% lock_players_by_id(min2)%>% set_player_min_exp(id= min3, exp= .2)%>% set_player_min_exp(id= min4, exp= .4)%>% set_player_min_exp(id= min5, exp= .4)

lineups2 <- build_lineups(opt, num_lineups = 150) summary(lineups2)`

anthonyshook commented 2 years ago

I can't reproduce the problem -- I think your example has an error in it (you're running build_lineups on opt, but defining your constraints on dkopt.

Here's what I ran -- it's identical to your example, except the lineups are using dkopt to build:

dkopt <- create_optimizer(site = 'DraftKings',
                          sport = 'BASKETBALL',
                          contest_type = 'CLASSIC')
df <- read.csv('./df.csv')
dkopt <- add_players_from_df(object = dkopt, df = df)

min <- get_player_id(dkopt,'Cam Reddish') #SG/SF
min2 <-get_player_id(dkopt,'Willy Hernangomez') #PF/C
min3 <-get_player_id(dkopt,'Delon Wright') #PG
min4 <-get_player_id(dkopt,'Saddiq Bey') #SF/SG
min5 <-get_player_id(dkopt,'LeBron James') #SF/PF

dkopt<-dkopt %>%
  lock_players_by_id(min) %>%
  lock_players_by_id(min2) %>%
  set_player_min_exp(id= min3, exp= .2) %>%
  set_player_min_exp(id= min4, exp= .4) %>%
  set_player_min_exp(id= min5, exp= .4)

lineups2 <- build_lineups(dkopt, num_lineups = 150)
summary(lineups2)

And here's the output

Number of Expected Lineups: 150
Number of Lineups Found: 150

Player Exposures:
                      Player Lineups     Percent
 1:              Cam Reddish     150 100.0000000
 2:        Willy Hernangomez     150 100.0000000
 3:              Tyler Herro     136  90.6666667
 4:               Saddiq Bey     102  68.0000000
 5:            Jalen Brunson      82  54.6666667
 6:               Seth Curry      78  52.0000000
 7:            Deandre Ayton      69  46.0000000
 8:             LeBron James      60  40.0000000
 9:             Franz Wagner      44  29.3333333
10:             Bradley Beal      43  28.6666667
11:           Dewayne Dedmon      41  27.3333333
12:             Delon Wright      30  20.0000000
13:           Brandon Ingram      30  20.0000000
14:                Josh Hart      29  19.3333333
15:             JaVale McGee      27  18.0000000
16:          Duncan Robinson      15  10.0000000
17:           Hamidou Diallo      14   9.3333333
18:             Myles Turner      13   8.6666667
19:             Monte Morris      10   6.6666667
20:            Julius Randle       9   6.0000000
21:             Gabe Vincent       7   4.6666667
22:          Jordan Clarkson       7   4.6666667
23:           Keldon Johnson       6   4.0000000
24:                Max Strus       6   4.0000000
25:               Kyle Lowry       5   3.3333333
26:             Desmond Bane       4   2.6666667
27:            Kyle Anderson       4   2.6666667
28:               Tyus Jones       3   2.0000000
29:             Caris LeVert       3   2.0000000
30:            Tobias Harris       3   2.0000000
31:            Mikal Bridges       3   2.0000000
32:            Malik Beasley       3   2.0000000
33:      Dorian Finney-Smith       2   1.3333333
34:              Joel Embiid       2   1.3333333
35:             Aaron Gordon       1   0.6666667
36:             Steven Adams       1   0.6666667
37:            Cameron Payne       1   0.6666667
38:         Bojan Bogdanovic       1   0.6666667
39:           Onyeka Okongwu       1   0.6666667
40:              Gary Harris       1   0.6666667
41: Nickeil Alexander-Walker       1   0.6666667
42:             Devin Booker       1   0.6666667
43:             Kevon Looney       1   0.6666667
44:         Donovan Mitchell       1   0.6666667
                      Player Lineups     Percent

Players Per Team:
ATL DAL DEN DET  GS IND LAL MEM MIA MIN  NO  NY ORL PHI PHO  SA UTA WAS 
181  84  11 116   1  16  60  12 210   3 210   9  45  83 101   6   9  43 

Lineup Variance (mean Jaccard Distance): 58.11 %