高校数学の質問スレ（医者・東大卒専用） Part438

高校数学の質問スレ（医者・東大卒専用） Part438 (979ﾚｽ)
上下前次1-新
通常表示 512ﾊﾞｲﾄ分割ﾚｽ栞
抽出解除ﾚｽ栞

113: １３２人目の素数さん [sage] 2024/08/21(水) 11:44:00.03 ID:trPnwZW4

シミュレーションによる検証

Wolfram Language 14.0.0 Engine for Microsoft Windows (64-bit)
Copyright 1988-2023 Wolfram Research, Inc.

In[1]:= j[n_] :=( (* n人でジャンケンして勝者が決まるまでの回数と勝者の数*)
count=0;
Until[Length@Union@a==2,a=RandomChoice[Range[3],n];count++];
b=Sort@Union@a;
If[b=={1,2}, winners=Count[a,2]];
If[b=={2,3}, winners=Count[a,3]];
If[b=={1,3}, winners=Count[a,1]] ;
{winners,count}
)

In[2]:=
In[2]:= sim[n_] :=((* 勝者が一人になるまでの回数　*)
For[{winner,counts}=j[n],winner>1,k=j[winner];winner=k[[1]];counts=counts+k[[2]]];
counts
)

In[3]:= res11=Table[sim[11],10^6];

In[4]:= Histogram[res11,"Scott","PDF"]

Out[4]= -Graphics-

In[5]:= Mean[res11] // N

Out[5]= 34.9504

In[6]:= Median[res11]

Out[6]= 27

In[7]:= N@Mean@Boole[#<=10&/@res11]

Out[7]= 0.154785

In[8]:= 449687340186660888579056289638229806808082/2909321189362570808630465826492242446680483

449687340186660888579056289638229806808082
Out[8]= -------------------------------------------
2909321189362570808630465826492242446680483

In[9]:= % // N

Out[9]= 0.154568

http://rio2016.5ch.net/test/read.cgi/math/1723152147/113

794: １３２人目の素数さん [sage] 2025/04/23(水) 02:35:47.03 ID:t2PViPB2

# Candidate values for the number of red balls (0 to 100)
R_vals <- 0:100

# Observed data
k <- 4      # Number of red balls drawn
n <- 10     # Sample size
N <- 100    # Total number of balls

# Likelihood using the hypergeometric distribution
likelihood <- dhyper(k, R_vals, N - R_vals, n)

# Prior distribution: uniform
prior <- rep(1, length(R_vals))

# Unnormalized posterior
posterior_unnorm <- likelihood * prior

# Normalize to get the posterior distribution
posterior <- posterior_unnorm / sum(posterior_unnorm)

# MAP estimate (most probable value)
R_MAP <- R_vals[which.max(posterior)]

# Posterior mean (expected value)
R_mean <- sum(R_vals * posterior)

# 95% central credible interval
cumulative <- cumsum(posterior)
lower_CI <- R_vals[which(cumulative >= 0.025)[1]]
upper_CI <- R_vals[which(cumulative >= 0.975)[1]]

# 95% Highest Posterior Density Interval (HPDI)
sorted <- order(posterior, decreasing = TRUE)
cumsum_sorted <- cumsum(posterior[sorted])
HPDI_index <- sorted[which(cumsum_sorted <= 0.95)]
HPDI_range <- range(R_vals[HPDI_index])

# Display results
cat("MAP estimate:", R_MAP, "\n")
cat("Posterior mean:", round(R_mean, 2), "\n")
cat("95% central credible interval: [", lower_CI, ",", upper_CI, "]\n")
cat("95% HPDI: [", HPDI_range[1], ",", HPDI_range[2], "]\n")

# Plot posterior distribution
plot(R_vals, posterior, type = "h", lwd = 2,
main = "Posterior Distribution P(R | Data)",
xlab = "Number of Red Balls (R)", ylab = "Posterior Probability")
abline(v = c(lower_CI, upper_CI), col = "blue", lty = 2)
abline(v = HPDI_range, col = "red", lty = 3)
legend("topright", legend = c("95% Central CI", "95% HPDI"),
col = c("blue", "red"), lty = c(2,3))

http://rio2016.5ch.net/test/read.cgi/math/1723152147/794

上下前次1-新書関写板覧索設栞歴

ｽﾚ情報赤ﾚｽ抽出画像ﾚｽ抽出歴の未読ｽﾚ AAｻﾑﾈｲﾙ

ぬこの手ぬこTOP 0.042s