#探索的因子分析
#ドラマ特性
x <- read.delim("clipboard")
install.packages(c("lavaan","semTools","semPlot")) 
library(lavaan)
library(semTools)
library(semPlot)
factanal(x,factors=5,rotation="promax")

#動機要因　#消費者特性
x <- read.delim("clipboard")
install.packages(c("lavaan","semTools","semPlot")) 
library(lavaan)
library(semTools)
library(semPlot)
factanal(x,factors=8,rotation="promax")


#確認的因子分析
HS.model.cfa <- '
akogare=~akogare1+akogare2+akogare3
zenwakanketu=~zenwakanketu1+zenwakanketu2+zenwakanketu3
gensaku=~gensaku1+gensaku2+gensaku3
torikata=~torikata1+torikata2+torikata3'

HS.model.cfa <- '
sukinahaiyu=~sukinahaiyu1+sukinahaiyu2+sukinahaiyu3
koukoku=~koukoku1+koukoku2+koukoku3
kutikomi=~kutikomi1+kutikomi2+kutikomi3
tyomei=~tyomei1+tyomei2+tyomei3
gisikiteki=~gisikiteki1+gisikiteki2+gisikiteki3
syudanteki=~syudanteki2+syudanteki3
zenkai=~zenkai1+zenkai2+zenkai3
nitiyou=~nitiyou1+nitiyou2+nitiyou3'

HS.model.cfa <- '
syakousei=~syakousei1+syakousei2+syakousei3+syakousei4
ryukou=~ryukou1+ryukou2+ryukou3
syukanteki=~syukanteki1+syukanteki2+syukanteki3
jikokenji=~jikokenji1+jikokenji2+jikokenji4
kaigaidorama=~kaigaidorama1+kaigaidorama2+kaigaidorama3
yokanosukunasa=~yokanosukunasa1+yokanosukunasa2+yokanosukunasa3
mame=~mame1+mame2+mame3'

fit <- lavaan(HS.model.cfa, data=x,　 auto.var=TRUE, auto.fix.first=TRUE,auto.cov.lv.x=TRUE)
summary(fit, fit.measures=TRUE,standardized=T,rsquare=T)	
modindices(fit,sort=T)



#共分散構造分析
HS.model.sem <- '
sityouito=~sityouito1+sityouito2+sityouito3+sityouito4+sityouito5
sityouito ~akogare + zenwakanketu + gensaku + torikata
akogare=~akogare1+akogare2+akogare3
zenwakanketu=~zenwakanketu1+zenwakanketu2+zenwakanketu3
gensaku=~gensaku1+gensaku2+gensaku3
torikata=~torikata1+torikata2+torikata3'

HS.model.sem <- '
sityouito=~sityouito1+sityouito2+sityouito3+sityouito4+sityouito5
sityouito ~ sukinahaiyu+ koukoku + kutikomi + tyomei + gisikiteki + syudanteki + zenkai + nitiyou
sukinahaiyu=~sukinahaiyu1+sukinahaiyu2+sukinahaiyu3
koukoku=~koukoku1+koukoku2+koukoku3
kutikomi=~kutikomi1+kutikomi2+kutikomi3
tyomei=~tyomei1+tyomei2+tyomei3
gisikiteki=~gisikiteki1+gisikiteki2+gisikiteki3
syudanteki=~syudanteki2+syudanteki3
zenkai=~zenkai1+zenkai2+zenkai3
nitiyou=~nitiyou1+nitiyou2+nitiyou3'

HS.model.sem <- '
sityouito=~sityouito1+sityouito2+sityouito3+sityouito4+sityouito5
sityouito ~syakousei + ryukou + syukanteki + jikokenji + kaigaidorama + yokanosukunasa + mame
syakousei=~syakousei1+syakousei2+syakousei3+syakousei4
ryukou=~ryukou1+ryukou2+ryukou3
syukanteki=~syukanteki1+syukanteki2+syukanteki3
jikokenji=~jikokenji1+jikokenji2+jikokenji4
kaigaidorama=~kaigaidorama1+kaigaidorama2+kaigaidorama3
yokanosukunasa=~yokanosukunasa1+yokanosukunasa2+yokanosukunasa3
mame=~mame1+mame2+mame3'


fit.sem <- lavaan(HS.model.sem, data=x,　 auto.var=TRUE, auto.fix.first=TRUE,auto.cov.lv.x=TRUE)
summary(fit.sem, fit.measures=TRUE,standardized=T,rsquare=T)


#追加分析
#共分散構造分析
HS.model.sem <- '
sityouito=~sityouito1+sityouito2+sityouito3+sityouito4+sityouito5
sityouito ~seibetu + club+ doubleschool + arubaito+ keiji+ renai+ comedy+ history+ iryou+ action+ mistery+ gakuen+ family+ sf+ sonota'

fit.sem <- lavaan(HS.model.sem, data=x,  auto.var=TRUE, auto.fix.first=TRUE,auto.cov.lv.x=TRUE)
summary(fit.sem, fit.measures=TRUE,standardized=T,rsquare=T)


#二項ロジット分析
x <- read.delim("clipboard")
install.packages(c("lavaan","semTools","semPlot")) 
library(lavaan)
library(semTools)
library(semPlot)
res1<-glm(sityouhindo1~tvweek+tvday+tvdoramayear,data=x)
        summary(res1)

res1<-glm(sityouhindo1+sityouhindo2+sityouhindo3~sityouhonsuu+ seibetu + club+ doubleschool + arubaito+ keiji+ renai+ comedy+ history+ iryou+ action+ mistery+ gakuen+ family+ sf+ sonota,data=x)
        summary(res1)

#二項ロジット分析の相関係数
x <- read.delim("clipboard")
quant1<-cbind(x[,1:3])
quant2<-cbind(tvweek,tvday,tvdoramayear)
round(cor(quant1),3)


#クラスター分析
install.packages("dplyr")
x<- read.csv("a.csv", header=T)
kmeans_result<-kmeans(x[,1:11],8)
kmeans_result
clustered_data <- data.frame(x,kmeans_result$cluster) 
#平均求める
library(dplyr)
cluster_summary <- clustered_data %>%
+  group_by(kmeans_result.cluster) %>%
+     summarise(
+         sex_mean = mean(sex),
+         tvweek_mean = mean(tvweek),
+         tvday_mean = mean(tvday),
+         tvdoramayear_mean = min(tvdoramayear)
+     ) %>%
+     as.data.frame
cluster_summary