Rcmdr.kr: An R Commander User in Korea

통계 > 적합성 모델 > 다항 로짓 모델...
Statistics > Fit models > Multinomial logit model...

MASS 패키지에 있는 birthwt 데이터셋을 활용해보자. 먼저 birthwt 데이터셋을 활성화 시킨다. https://rcmdr.kr/149

이후 일부 변수의 특징을 변화시켜 bwt라는 새로운 데이터셋을 만들자.아래의 <일반화 선형 모델> 연습 중간에 bwt를 만드는 스크립트가 있다. https://rcmdr.tistory.com/150

아래 화면은 bwt 데이터셋을 활용한 사례이다. low 변수를 반응변수로, 다른 나머지 변수를 설명변수군에 모두 포함시키자.

MLM.1 <- multinom(low ~ ., data=bwt, trace=FALSE)
summary(MLM.1, cor=FALSE, Wald=TRUE)

MLM.1 이라는 모형이 만들어졌다면, '모델 > 가설 검정 > 분산분석표...' 메뉴 기능을 이용할 수 있다.

통계 > 적합성 모델 > 일반화 선형 모델...
Statistics > Fit models > Generalized linear model...

MASS 패키지에 있는 birthwt 데이터셋을 활용하자. https://rcmdr.tistory.com/149

birthwt 데이터셋을 <일반화 선형 모델> 분석을 위하여 변형시켜서 bwt라는 데이터셋으로 만들자.

bwt <- with(birthwt, {
race <- factor(race, labels = c("white", "black", "other"))
ptd <- factor(ptl > 0)
ftv <- factor(ftv)
levels(ftv)[-(1:2)] <- "2+"
data.frame(low = factor(low), age, lwt, race, smoke = (smoke > 0),
           ptd, ht = (ht > 0), ui = (ui > 0), ftv)
})

options(contrasts = c("contr.treatment", "contr.poly"))

bwt 데이터셋을 활성화시키자.

GLM.3 <- glm(low ~ ., family=binomial(logit), data=bwt)
summary(GLM.3)

exp(coef(GLM.3))  # Exponentiated coefficients ("odds ratios")

## an example with offsets from Venables & Ripley (2002, p.189)
utils::data(anorexia, package = "MASS")

GLM.2 <- glm(Postwt ~ Prewt + Treat + offset(Prewt), 
	family=gaussian(identity), data=anorexia)
summary(GLM.2)

data(Cowles, package="carData")

GLM.1 <- glm(volunteer ~ sex + neuroticism*extraversion, family=binomial(logit), data=Cowles)
summary(GLM.1)
exp(coef(GLM.1))  # Exponentiated coefficients ("odds ratios")

MASS::birthwt

data(birthwt, package="MASS")

birthwt 데이터셋이 활성화된 후, <데이터셋 보기> 버튼을 누르면 아래와 같이 내부 구성을 볼 수 있다:

help("birthwt")

Risk Factors Associated with Low Infant Birth Weight

Description

The birthwt data frame has 189 rows and 10 columns. The data were collected at Baystate Medical Center, Springfield, Mass during 1986.

Usage

birthwt

Format

This data frame contains the following columns:

low

indicator of birth weight less than 2.5 kg.

age

mother's age in years.

lwt

mother's weight in pounds at last menstrual period.

race

mother's race (1 = white, 2 = black, 3 = other).

smoke

smoking status during pregnancy.

ptl

number of previous premature labours.

ht

history of hypertension.

ui

presence of uterine irritability.

ftv

number of physician visits during the first trimester.

bwt

birth weight in grams.

Source

Hosmer, D.W. and Lemeshow, S. (1989) Applied Logistic Regression. New York: Wiley

References

Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.

Examples

bwt <- with(birthwt, {
race <- factor(race, labels = c("white", "black", "other"))
ptd <- factor(ptl > 0)
ftv <- factor(ftv)
levels(ftv)[-(1:2)] <- "2+"
data.frame(low = factor(low), age, lwt, race, smoke = (smoke > 0),
           ptd, ht = (ht > 0), ui = (ui > 0), ftv)
})
options(contrasts = c("contr.treatment", "contr.poly"))
glm(low ~ ., binomial, bwt)

모델 > 가설 검정 > 분산분석표...

Models > Hypothesis tests > ANOVA table...

Statistics > Fit models > Linear model... (통계 > 적합성 모델 > 선형 모델...) 기능을 통하여 LinearModel.1라는 모델을 만들었다고 하자. 위의 화면 상단에 <모델: LinearModel.1>이라고 보일 것이다. 

https://rcmdr.tistory.com/131

Models > Hypothesis tests > ANOVA table... (모델 > 가설 검정 > 분산분산표...) 기능은 앞선 모델 만들기 작업 위에서 후속적으로 가능함을 명심하자.

Anova(LinearModel.1, type="II")

Rcmdr > etc > Nations.txt (컴파일 이후)

Rcmdr > inst > etc > Nations.txt (소스에서)

Rcmdr 패키지에는 Getting-Started-with-the-Rcmdr.pdf 라는 사용법 소개 파일이 포함되어 있다. 이 파일에서 예제용으로 활용하는 데이터셋이 Nations.txt 이다. 이 파일 또한 Rcmdr 패키지에 포함되어 있다.

아래와 같이 내부 구성을 볼 수 있다.

Nations 데이터셋의 요약 정보는 아래와 같다:

TFR (출생율, the total fertility rate, expressed as number of children per woman),

contraception (피임율, the rate of contraceptive use among married women, in percent),

infant.mortality (영유아 사망율, the infant-mortality rate per 1000 live births),

GDP (국내 총생산, gross domestic product per capita, in U.S. dollars),

region (대륙, one line per country)

통계 > 차원 분석 > 요인 분석...

Statistics > Dimensional analysis > factor analysis...

통계 > 차원 분석 > 주-성분 분석...

Statistics > Dimensional analysis > Principal-components analysis...

<주성분 분석> 메뉴 창에서 <변수 (두개 이상 선택)> 에서 4개의 변수를 모두 선택해보자.

<선택기능> 창에서 기본 설정되어 있는 기능을 기억하자.

local({
  .PC <- princomp(~Assault+Murder+Rape+UrbanPop, cor=TRUE, data=USArrests)
  cat("\nComponent loadings:\n")
  print(unclass(loadings(.PC)))
  cat("\nComponent variances:\n")
  print(.PC$sd^2)
  cat("\n")
  print(summary(.PC))
})

.PC <- princomp(~Assault+Murder+Rape+UrbanPop, cor=TRUE, data=USArrests)
plot(.PC)

biplot(.PC)

datasets::USArrests()

data(USArrests, package="datasets")

R Commander 화면 상단에서 <데이터셋 보기> 버튼을 누르면 아래와 같은 내부 구성을 확인할 수 있다.

help("USArrests")

Violent Crime Rates by US State

Description

This data set contains statistics, in arrests per 100,000 residents for assault, murder, and rape in each of the 50 US states in 1973. Also given is the percent of the population living in urban areas.

Usage

USArrests

Format

A data frame with 50 observations on 4 variables.

Note

USArrests contains the data as in McNeil's monograph. For the UrbanPop percentages, a review of the table (No. 21) in the Statistical Abstracts 1975 reveals a transcription error for Maryland (and that McNeil used the same “round to even” rule that R's round() uses), as found by Daniel S Coven (Arizona).

See the example below on how to correct the error and improve accuracy for the ‘<n>.5’ percentages.

Source

World Almanac and Book of facts 1975. (Crime rates).

Statistical Abstracts of the United States 1975, p.20, (Urban rates), possibly available as https://books.google.ch/books?id=zl9qAAAAMAAJ&pg=PA20.

References

McNeil, D. R. (1977) Interactive Data Analysis. New York: Wiley.

See Also

The state data sets.

Examples

summary(USArrests)

require(graphics)
pairs(USArrests, panel = panel.smooth, main = "USArrests data")

## Difference between 'USArrests' and its correction
USArrests["Maryland", "UrbanPop"] # 67 -- the transcription error
UA.C <- USArrests
UA.C["Maryland", "UrbanPop"] <- 76.6

## also +/- 0.5 to restore the original  <n>.5  percentages
s5u <- c("Colorado", "Florida", "Mississippi", "Wyoming")
s5d <- c("Nebraska", "Pennsylvania")
UA.C[s5u, "UrbanPop"] <- UA.C[s5u, "UrbanPop"] + 0.5
UA.C[s5d, "UrbanPop"] <- UA.C[s5d, "UrbanPop"] - 0.5

## ==> UA.C  is now a *C*orrected version of  USArrests

carData > Prestige

data(Prestige, package="carData")

help("Prestige")

Prestige of Canadian Occupations

Description

The Prestige data frame has 102 rows and 6 columns. The observations are occupations.

Usage

Prestige

Format

This data frame contains the following columns:

education

Average education of occupational incumbents, years, in 1971.

income

Average income of incumbents, dollars, in 1971.

women

Percentage of incumbents who are women.

prestige

Pineo-Porter prestige score for occupation, from a social survey conducted in the mid-1960s.

census

Canadian Census occupational code.

type

Type of occupation. A factor with levels (note: out of order): bc, Blue Collar; prof, Professional, Managerial, and Technical; wc, White Collar.

Source

Canada (1971) Census of Canada. Vol. 3, Part 6. Statistics Canada [pp. 19-1–19-21].

Personal communication from B. Blishen, W. Carroll, and C. Moore, Departments of Sociology, York University and University of Victoria.

References

Fox, J. (2016) Applied Regression Analysis and Generalized Linear Models, Third Edition. Sage.

Fox, J. and Weisberg, S. (2019) An R Companion to Applied Regression, Third Edition, Sage.

carData > Moore

data(Moore, package="carData")

help("Moore")

Status, Authoritarianism, and Conformity

Description

The Moore data frame has 45 rows and 4 columns. The data are for subjects in a social-psychological experiment, who were faced with manipulated disagreement from a partner of either of low or high status. The subjects could either conform to the partner's judgment or stick with their own judgment.

Usage

Moore

Format

This data frame contains the following columns:

partner.status

Partner's status. A factor with levels: high, low.

conformity

Number of conforming responses in 40 critical trials.

fcategory

F-Scale Categorized. A factor with levels (note levels out of order): high, low, medium.

fscore

Authoritarianism: F-Scale score.

Source

Moore, J. C., Jr. and Krupat, E. (1971) Relationship between source status, authoritarianism and conformity in a social setting. Sociometry 34, 122–134.

Personal communication from J. Moore, Department of Sociology, York University.

References

Fox, J. (2016) Applied Regression Analysis and Generalized Linear Models, Third Edition. Sage.

Fox, J. and Weisberg, S. (2019) An R Companion to Applied Regression, Third Edition, Sage.