Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
R を用いた検定(補講) (1) — Welch 検定 / Tests using R (su...
Search
Kenji Saito
PRO
November 30, 2024
Technology
0
51
R を用いた検定(補講) (1) — Welch 検定 / Tests using R (supplementary) (1) - Welch test
早稲田大学大学院経営管理研究科「企業データ分析」2024 冬のオンデマンド教材 第9回で使用したスライドです。
Kenji Saito
PRO
November 30, 2024
Tweet
Share
More Decks by Kenji Saito
See All by Kenji Saito
マニフェスト: 人類の知のフロンティアに向けた拡張的足場へ / Manifesto: Toward Expansive Scaffolding for Humanity's Knowledge Frontier
ks91
PRO
0
8
続・スマートコントラクトと分散ファイナンス / Smart Contracts and Decentralized Finance, Continued
ks91
PRO
0
43
スマートコントラクトと分散ファイナンス / Smart Contracts and Decentralized Finance
ks91
PRO
0
63
シン・ブロックチェーン / Truth of Blockchain
ks91
PRO
0
97
パスワード/パスフレーズと認証 / Password, Passphrase and Authentication
ks91
PRO
0
37
git と GitHub / git and GitHub
ks91
PRO
0
36
ソフトウェアの開発と保守 / Software Development and Maintenance
ks91
PRO
0
50
インターネットの特徴 / Features of the Internet
ks91
PRO
0
34
インターネットのガバナンス / Governance of the Internet
ks91
PRO
0
30
Other Decks in Technology
See All in Technology
AWS認定を取る中で感じたこと
siromi
1
190
Glacierだからってコストあきらめてない? / JAWS Meet Glacier Cost
taishin
1
170
Delta airlines Customer®️ USA Contact Numbers: Complete 2025 Support Guide
deltahelp
0
820
データ基盤からデータベースまで?広がるユースケースのDatabricksについて教えるよ!
akuwano
3
130
DBのスキルで生き残る技術 - AI時代におけるテーブル設計の勘所
soudai
PRO
54
20k
成長し続けるアプリのためのテストと設計の関係、そして意思決定の記録。
sansantech
PRO
0
130
スタートアップに選択肢を 〜生成AIを活用したセカンダリー事業への挑戦〜
nstock
0
250
改めてAWS WAFを振り返る~業務で使うためのポイント~
masakiokuda
2
270
Lufthansa ®️ USA Contact Numbers: Complete 2025 Support Guide
lufthanahelpsupport
0
210
United Airlines Customer Service– Call 1-833-341-3142 Now!
airhelp
0
170
american airlines®️ USA Contact Numbers: Complete 2025 Support Guide
supportflight
1
110
Contributing to Rails? Start with the Gems You Already Use
yahonda
2
100
Featured
See All Featured
Statistics for Hackers
jakevdp
799
220k
Cheating the UX When There Is Nothing More to Optimize - PixelPioneers
stephaniewalter
281
13k
Rebuilding a faster, lazier Slack
samanthasiow
83
9.1k
Fantastic passwords and where to find them - at NoRuKo
philnash
51
3.3k
10 Git Anti Patterns You Should be Aware of
lemiorhan
PRO
656
60k
How to Think Like a Performance Engineer
csswizardry
25
1.7k
How GitHub (no longer) Works
holman
314
140k
Measuring & Analyzing Core Web Vitals
bluesmoon
7
510
Building a Modern Day E-commerce SEO Strategy
aleyda
42
7.4k
Typedesign – Prime Four
hannesfritz
42
2.7k
Refactoring Trust on Your Teams (GOTO; Chicago 2020)
rmw
34
3.1k
Building Applications with DynamoDB
mza
95
6.5k
Transcript
Boxes and whiskers — generated by Stable Diffusion XL v1.0
2024 9 R ( ) (1) — Welch (WBS) 2024 9 R ( ) (1) — Welch — 2024-11 – p.1/10
https://speakerdeck.com/ks91/collections/corporate-data-analysis-2024-winter 2024 9 R ( ) (1) — Welch —
2024-11 – p.2/10
( 20 ) 1 • 2 R • 3 •
4 • 5 • 6 ( ) • 7 (1) • 8 (2) • 9 R ( ) (1) — Welch • 10 R ( ) (2) — χ2 11 R ( ) (1) — 12 R ( ) (2) — 13 GPT-4 14 GPT-4 15 ( ) LaTeX Overleaf 8 (12/16 ) / (2 ) OK / 2024 9 R ( ) (1) — Welch — 2024-11 – p.3/10
t Welch R t.test() Welch Welch 2024 9 R (
) (1) — Welch — 2024-11 – p.4/10
2 t ( ) (1/2) 2 ( ) xA −
xB (1) : (2) : σ ( ) σ sp sp = s2 A (nA − 1) + s2 B (nB − 1) nA + nB − 2 (R var() ) nA + nB − 2 t Welch A B (µA = µB ) A B (µA = µB ) 2024 9 R ( ) (1) — Welch — 2024-11 – p.5/10
2 t ( ) (2/2) xA − xB Student µA
= µB t = (xA − xB ) − (µA − µB ) sp 1 nA + 1 nB = xA − xB sp 1 nA + 1 nB (t ) t dfp = nA + nB − 2 t ( ) t0.05 (dfp ) t0.05 (dfp ) < |t| (P < 0.05) 2024 9 R ( ) (1) — Welch — 2024-11 – p.6/10
Welch t t = xA − xB s2 A nA
+ s2 B nB ( ) v . . . v ≈ ( s2 A nA + s2 B nB )2 s4 A n2 A (nA−1) + s4 B n2 B (nB−1) R 2024 9 R ( ) (1) — Welch — 2024-11 – p.7/10
( ) - (Shapiro-Wilk test) - (Anderson-Darling test for normality)
- (Kolmogorov-Smirnov test for normality) ( ) ( ) (Bartlett’s test for homogeneity of variances) 2024 9 R ( ) (1) — Welch — 2024-11 – p.8/10
.txt A /B g <- read.table(" .txt", header=T) colnames(g) <-
c(" ", " ") sampleA <- g$ sampleB <- g$ # ( ) shapiro.test(x=sampleA) shapiro.test(x=sampleB) # ( ) samples <- c(sampleA, sampleB) group_factor <- factor(rep(c("A", "B"), c(length(sampleA), length(sampleB)))) bartlett.test(formula=samples~group_factor) # Welch (t.test() ) ( Welch ) t.test(sampleA, sampleB) 2024 9 R ( ) (1) — Welch — 2024-11 – p.9/10
2024 9 R ( ) (1) — Welch — 2024-11
– p.10/10