library(plyr)
library(stats)
library(ggplot2)
library(moments)
library(psych)
library(car)
library(fitdistrplus)

data("Prestige")

stat<-as.data.frame(matrix(c(

mean(Prestige$prestige),
length(Prestige$prestige),
min(Prestige$prestige),
max(Prestige$prestige) ,
range=max(Prestige$prestige)-min(Prestige$prestige) ,
sd(Prestige$prestige) ,
median(Prestige$prestige),
quantile(Prestige$prestige, probs=0.25),
quantile(Prestige$prestige, probs=0.75),
skewness(Prestige$prestige),
kurtosis(Prestige$prestige),
IQR(Prestige$prestige),
mean(Prestige$prestige, trim=0.05),
mean(Prestige$prestige, trim=0.1),
winsor.mean(Prestige$prestige, trim=0.05),
winsor.mean(Prestige$prestige, trim=0.1)
), nrow=1)
)

colnames(stat)<-c("mean", "length", "min", "max", "range", "sd", "median", "q25", "q75", "skewness", "kurtosis", "IQR",
                   "mean 0.05", "mean 0.1", "wins 0.05", "wins 0.1")



#dopasowanie rozkladu

qplot(Prestige$prestige, geom="histogram", xlab = "prestige")


plot(density(Prestige$prestige), main ="Gęstość zmiennej prestige")


stat2<-fitdist(Prestige$prestige, "norm")

plot(fitdist(Prestige$prestige, "norm", "mle"))

plot(fitdist(Prestige$prestige, "norm","mge"))

cor(Prestige$prestige, Prestige$education)
cor(Prestige$prestige, Prestige$income)


ks.test(Prestige$prestige, "pnorm", mean=46.83333, sd=17.11994)



limits<-c(0, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100)
table<-classIntervals(Prestige$prestige,n=9, style="fixed", fixedBreaks=limits, intervalClosure = 'right')


plot(table, pal=rainbow(length(limits) - 1), main='Prestige', xlab='Współczynnik prestiżu', ylab='procent zawodów')


jenks.tests(table)



#-----------regresja----------------------------------------------------------------------


scatterplot(x=Prestige$income, y=Prestige$prestige, main="Income ~ Education")


scatter.smooth(x=Prestige$income, y=Prestige$education, main="Income ~ Education")