所以我使用这些线来绘制我的强盗策略,但它们都具有完全相同的累积奖励率。这是不可能的。我在这里做错了什么?提前致谢!:)
simulations <- 1
horizon <- nrow(atukunda_a) #My dataset is atukunda_a
log_S <- atukunda_a #This is the data
formula <- "blood_loss_24hrs ~ study_group" #reward ~ arms
bandit <- OfflineReplayEvaluatorBandit$new(formula,log_S)
#?RandomPolicy
agents <-
list(Agent$new(ThompsonSamplingPolicy$new(1.0,1.0), bandit, "ThompsonSampling"),
Agent$new(UCB1Policy$new(), bandit, "UCB1"))
simulation <-
Simulator$new(
agents = agents,
simulations = simulations,
horizon = horizon,
save_context = FALSE,
progress_file = TRUE,
do_parallel = FALSE
)
sim <- simulation$run()
plot(sim, type = "cumulative", legend_title = "Simulation",
rate = TRUE, regret = FALSE, legend_position = "bottomright")