# source("E:\\mini_curso_R_CESPE\\Script.txt")

########################################################
########################################################
########################################################
########################################################
########################################################
#
# Operações básicas, vetores e matrizes
#
########################################################
########################################################
########################################################
########################################################
########################################################

######################
# Somando dois números

2+2

######################
# Subtraindo-se dois números

2-2

######################
# Multiplicando-se dois números

2*2

######################
# Dividindo-se dois números

2/2

##########################
# Definindo um vetor linha

x <- cbind(1,3,7,5,2)

x


##########################
# Definindo um vetor coluna

y <- rbind(15,20,28,21,18)

y


##########################
# Transformando o vetor x em vetor coluna

x <- t(x)

x


# Funçoes no R
# Tem como entrada matrizes e como saída uma lista de objetos


# help : ?nomedafunção

# help.search("nomedafuncao")

########################################################
########################################################
########################################################
########################################################
########################################################
#
# Gráficos e algumas funções
#
########################################################
########################################################
########################################################
########################################################
########################################################




##########################
# Calculando-se esatísticas descritivas de cada variável

summary(x)

summary(y)


# media, variância, desvio-padrão e coeficiente de variação


cat("A média da variável x é :",mean(x),"\n")

cat("A variância da variável x é :",var(x),"\n")

cat("O desvio-padrão da variável x é :",sqrt(var(x)),"\n")

cat("O coeficiente de variação da variável x é :",(sqrt(var(x))/mean(x))*100,"\n")


# Criando uma matriz com as estatítisticas descritivas de x e y


est.desc.xy <- cbind(rbind(mean(x),var(x),sqrt(var(x)),(sqrt(var(x))/mean(x))*100),
rbind(mean(y),var(y),sqrt(var(y)),(sqrt(var(y))/mean(y))*100))


est.desc.xy 


est.desc.xy.1 <- round(est.desc.xy,2)


est.desc.xy.1

 

# Criando um tabela de medidas descritivas para os valores


rbind(summary(x),paste("Var.   :",var(x)),paste("DP    : ",round(sqrt(var(x)),1)),paste("CV(%)   : ",round((sqrt(var(x))/mean(x))*100,1)))


#############################
#
# Histogramas


hist(x)


hist(y)


# Adicionando-se informações no gráfico

hist(x,xlab="nota", probability=TRUE,ylab="densidade",main="Histograma da nota")


hist(y,xlab="renda",probability=TRUE, ylab="densidade",main="Histograma da renda")


##########################
# Plotando os valores de y versus x

plot(x,y)



##########################
# Mais detalhes no gráfico

plot(x,y,xlab="renda em salarios minimos dos pais",ylab="escore do respondente",main="Renda dos pais x Nota")
 



############################
#
# Box-plot


boxplot(x,main="Box-plot da var. renda")

boxplot(y,main="Box-plot das var. notas")




##########################
# Definindo-se uma matriz a partir dos vetores Calculando-se esatísticas descritivas de cada variável

z <- cbind(x,y)


############################
# Medidas resumo da matriz z


summary(z)


#############################
# Operações matriciais


# Multiplicação elemento a elemento

z*z

# Multiplicação matrical

w <- t(z)%*%z

w

# Inversão

solve(w)


# Produto de Kronecker


v<- kronecker(z,t(z))


# Referências a elementos de z

z[1,1]

z[,1]

z[2,]

z[,1:2]

z[1:2,]

z[1:2,1:2]


########################################################
########################################################
########################################################
########################################################
########################################################
#
# Listas e data.frame
#
########################################################
########################################################
########################################################
########################################################
########################################################



##################################
# Criando-se uma lista de objetos
#
# Útilo para se armazenar resultados
# de funções



ml <- list(renda=x,nota=y)


ml$renda

ml$nota



###################################
# Atribuindo-se nomes as observações


 nomes<-rbind("Caio","Girlene","Camila","Amanda","Gabriela")


 rownames(z) <- nomes


# rownames(ml) <- nomes não é possível utilizar tal recurso em uma lista


##########################
# Atribuindo-se nomes as variáveis


 variav<-rbind("renda","nota")


 colnames(z) <- variav


# colnames(ml) <- nomes não é possível utilizar tal recurso em uma lista

 


##########################################################
# Matrizes e listas possuem propriedas comuns e algumas
# diferenças. Seria possível unir todas as características
#
#
#


##########################################
# Definindo-se um data frame a partir de z 
#
# Data framE: matriz com propriedades de lista
#
#


z<- data.frame(z)

z[,1]

z[,2]

z$renda

z$nota


########################################################################
# Definindo-se mais duas variáveis (categorizadas) para o banco de dados

 formacao <- rbind("Est","Psi","Psi","Est","Est")

 proced <- rbind("Nordeste","Nordeste","Sudeste","Sudeste","Sudeste")


#######################################################################
# Adicionando-se essas variáveis ao data frame


z<- cbind(z,formacao,proced)


################################
# Listar os "objetos" disponíveis


ls()


#################################
# Salva o data.frame z


write.table(z,file="E:\\mini_curso_R_CESPE\\dadoscurso.txt")


##################################
# Remove todos os objetos da
# área de trabalho

rm(z)


ls()

z



#################################
# Carregando o data.frame z


z<-read.table(file="E:\\mini_curso_R_CESPE\\dadoscurso.txt")

z



########################################################
########################################################
########################################################
########################################################
########################################################
#
# Alguns métodos estatísticos
#
########################################################
########################################################
########################################################
########################################################
########################################################



######################################
# Ajustando um modelo de regressão
# para explicar nota em funcao da renda


# Lembrando


plot(x,y,xlab="renda em salarios minimos dos pais",ylab="escore do respondente",main="Renda dos pais x Nota")
 

resultreg<-lm(z$nota~z$renda)


resultreg


# Visualizando os resultados do ajuste do modelo

summary(resultreg)


# Verificando os objetos disponíveis

names(resultreg)

# Visualizando os coeficientes da regressão

resultreg$coefficients


# Armazenando os coeficientes num vetor (sem os nomes)

beta<-cbind(c(cbind(resultreg$coefficients)))




fitted <- resultreg$fitted.values


residual <- resultreg$residuals


# Plotando os valores observados versos os preditos



plot(z$renda,z$nota,xlab="renda em salarios minimos dos pais",ylab="escore do respondente",main="Renda dos pais x Nota",pch=19)

 

 
lines(z$renda,fitted,type="p",pch=17,col=3)



# Adicionando-se uma legenda ao gráfico


legend(2,26, legend=rbind("observado","ajustado"),pch=cbind(19,17),col=rbind(1,3))




# Adicionando-se a reta ajustada


m.X <- cbind(seq(1,1,length.out=100),seq(1,7,length.out=100))


plot(z$renda,z$nota,xlab="renda em salarios minimos dos pais",ylab="escore do respondente",main="Renda dos pais x Nota",pch=19)


 
lines(z$renda,fitted,type="p",pch=17,col=3)


lines(m.X[,2],m.X %*% beta,lty=2,col=4)



legend(2,26, legend=rbind("observado","ajustado","reta ajustada"),pch=cbind(19,17,46),lty=cbind(0,0,4),col=rbind(1,3,4))




# Colocar vários gráficos numa só Janela



par(mfrow=c(1,3))


# Valores observados, preditos e reta ajusatda




plot(z$renda,z$nota,xlab="renda em salarios minimos dos pais",ylab="escore do respondente",main="Renda dos pais x Nota",pch=19)


 
lines(z$renda,fitted,type="p",pch=17,col=3)


lines(m.X[,2],m.X %*% beta,lty=2,col=4)



legend(2,26, legend=rbind("observado","ajustado","reta ajustada"),pch=cbind(19,17,46),lty=cbind(0,0,4),col=rbind(1,3,4))





# Resíduos x renda

plot(z$renda,residual,xlab="renda dos pais",ylab="residuos",main="Resíduos x renda")

abline(0,0)



# Resíduos x valores preditos

plot(fitted,residual,xlab="valores preditos",ylab="residuos",main="Resíduos x valores preditos")

abline(0,0)





# Lista de parâmetros dos gráficos

  # ?par






##########################
# Tabelas de contingência


 # Variáveis categorizadas

  table(z$formacao,z$proced)


result.table <-table(z$formacao,z$proced)


summary(result.table)


names(result.table)


###############################
# Medidas descritivas pro grupo



tapply(z$nota,z$proced,mean)


tapply(z$nota,z$formacao,mean)


tapply(z$renda,z$proced,mean)


tapply(z$renda,z$formacao,mean)


tapply(z$nota,list(z$proced,z$formacao),mean)


################
#
# Box plot
#


boxplot(z$nota~z$proced)


boxplot(z$nota~z$formacao)


#boxplot(z$nota~(z$formacao*z$proced))



#######################
#######################
#######################
# Análise de variância


# Fator 1: formacao

formacao <- C(z$formacao)


# Fator 2: procedencia

proced <- C(z$proced)


#anova <- lm(z$nota ~ C(z$formacao,contr="treatment") + C(z$proced,contr="treatment"))



anova <- lm(z$nota ~ formacao + proced)

#

summary(anova)



##################################
#
# Análise de bancos de dados reais
#


#library(help="datasets")




# inteligence.data<-cbind(iris,rbind(x1,x2))

# x1<-cbind(rep("Publica",90))

# x2<-cbind(rep("Privada",60))

# Maximo por abilidade : 8, 6, 7, 3

# write.table(int.data,file="E:\\mini_curso_R_CESPE\\inteligence.data.txt")

int.data<-read.table(file="E:\\mini_curso_R_CESPE\\inteligence.data.txt")


cor(int.data[,1:4])


Espacial <- cbind(int.data$Espacial)


Linguistica<- cbind(int.data$Linguistica)


Interpretativa <- cbind(int.data$Interpretativa)


Logica <- cbind(int.data$Logica)



Social<- int.data$Social


Escola <-int.data$Escola




###################################
#
# Histogramas das variáveis
#


par(mfrow=c(2,2))


hist(int.data[int.data[,6]=="publica",1],xlab="nota",ylab="frequencia",probability=TRUE,main="Histograma: publica Espacial")


hist(int.data[int.data[,6]=="privada",1],xlab="nota",ylab="frequencia",probability=TRUE,main="Histograma: privada Espacial")


hist(int.data[int.data[,6]=="publica",2],xlab="nota",ylab="frequencia",probability=TRUE,main="Histograma: publica Linguistica")


hist(int.data[int.data[,6]=="privada",2],xlab="nota",ylab="frequencia",probability=TRUE,main="Histograma: privada Linguistica")



######################################
#
# Medidas resumo
#



tapply(Espacial,list(Social,Escola),mean)

tapply(Linguistica,list(Social,Escola),mean)

tapply(Interpretativa,list(Social,Escola),mean)

tapply(Logica,list(Social,Escola),mean)



#################################
#
# Box-plot por grupo para cada variável
#
#

par(mfrow=c(2,2))

# Social


boxplot(Espacial~Social ,main="Box-plot Espacial")


boxplot(Linguistica~Social ,main="Box-plot Linguistica")


boxplot(Interpretativa~Social  ,main="Box-plot Interpretativa ")


boxplot(Logica~Social ,main="Box-plot Logica")



par(mfrow=c(2,2))

# Escola



boxplot(Espacial~Escola,main="Box-plot Espacial")


boxplot(Linguistica~Escola ,main="Box-plot Linguistica")


boxplot(Interpretativa~Escola ,main="Box-plot Interpretativa ")


boxplot(Logica~Escola,main="Box-plot Logica")



par(mfrow=c(2,2))


# Escola x Social



boxplot(Espacial~Social*Escola,main="Box-plot Espacial")


boxplot(Linguistica~Social*Escola ,main="Box-plot Linguistica")


boxplot(Interpretativa~Social*Escola ,main="Box-plot Interpretativa ")


boxplot(Logica~Social*Escola,main="Box-plot Logica")



# Package Sciplot


# lineplot.CI





#############################
#############################
#
# Componentes principais
#


anacomp <- princomp(int.data[,1:4],cor=TRUE)


screeplot(anacomp)


anacomp


summary(anacomp)


escores.int <- cbind(-(anacomp $scores)[,1])



par(mfrow=c(2,2))



boxplot(escores.int~Social,main="Box-plot Espacial")


boxplot(escores.int~Escola,main="Box-plot Espacial")


boxplot(escores.int~Social*Escola,main="Box-plot Espacial")



# Gráficos de perfis médios


lineplot.CI(x.factor=Social,response=escore.int,group=Escola)


###############################
#
# Anova para O escore
#

sfac <- C(Social)


efac <- C(Escola)


anovaprin <- lm(escores.int ~sfac + efac)


summary(anovaprin)


#############################
#############################
#
# Análise fatorial
#



anafat <- factanal(int.data[,1:4],factors=1,scores=c("regression"))


anafat


summary(anafat)


names(anafat)


fac.int <- anafat$scores




par(mfrow=c(2,2))



boxplot(fac.int~Social,main="Box-plot Espacial")


boxplot(fac.int~Escola,main="Box-plot Espacial")


boxplot(fac.int~Social*Escola,main="Box-plot Espacial")





###############################
#
# Anova para O escore
#

sfac <- C(Social)


efac <- C(Escola)


anovaprin <- lm(fac.int ~sfac + efac)


summary(anovaprin)