#############################################################################
#                            ESTATISTICA DESCRITIVA                         #
#############################################################################

#### Setar diretorio ####

# Atalho -> ctrl + shift + h

# Indicando o caminho -> setwd("C:/Users/Downloads")

# Seta automaticamente o diretorio onde o arquivo esta salvo 
setwd(dirname(rstudioapi::getActiveDocumentContext()$path))

# verificar diretorio -> getwd()


#############################################################################
#### Entrando com os dados ####

# os dados devem estar na pasta indicado no diretorio
# neste caso, na pasta C:/Users/Downloads

# leitura de dados em arquivo externo
dados = read.table("dadosProgramaR.txt", dec = ",", header = T)   

# outra opcao para entrar com os dados:
# dados = read.table(file = file.choose(), dec = "," ,header = T)   


# Extrai as variaveis do objeto dados. Dessa forma, as variaveis do
# conjunto de dados podem ser acessadas simplesmente fornecendo seus nomes
attach(dados)
# Obs.: nao recomendado se for trabalhar com mais de um conjunto de 
# dados no mesmo arquivo R.


# Extrai os nomes das variaveis do objeto dados
names(dados)

# Extrai as seis primeiras linhas do conjunto de dados
head(dados)


#############################################################################
############################# DADOS QUALITATIVOS ############################

#### Tabela de frequencia para dados qualitativos ####
table(EsT_Civ)
table(G_Int)
table(EsT_Civ, G_Int)
table(EsT_Civ, G_Int, R_proced)

 
#### Tabela de proporcoes ####
prop.table(table(EsT_Civ))
prop.table(table(EsT_Civ, G_Int))


# Proporcao segundo linha
prop.table(table(EsT_Civ, G_Int), 1)


# Proporcao segundo coluna
prop.table(table(EsT_Civ, G_Int), 2)


#### Grafico de Barras ####
barplot(table(G_Int))

## personalizando ##
barplot(table(EsT_Civ), col = c("green", "red"), ylim = c(0, 25), 
        space = 0.8, width = c(0.2, 0.2),
        main = "Proporcao de funcionario por estado civil",
        xlab = "Estado Civil", ylab = "Proporcao de funcionarios")
# Adicionando as proporcoes
text(locator(n = 2), c("56%", "44%"))

# ou

barplot(table(EsT_Civ), ylim = c(0, 25), space = 0.8, width = c(0.2, 0.2), 
        col = c("green", "red"), legend = c("56%", "44%"), 
        main = "Proporcao de funcionario por estado civil",
        xlab = "Estado Civil", ylab = "Proporcao de funcionarios")



#### Grafico de Setores ou Pizza ####
pie(table(EsT_Civ), col = c("red", "yellow"), 
    labels = c("Casado (56%)", "Solteiro (44%)"),
    main = "Distribuicao dos Funcionarios por Estado Civil")


pie(table(G_Int), col = c("red", "yellow", "green"),
    labels = c("1 Grau (33,3%)", "2 Grau (50,0%)", "Superior (15,7%)"))
title(main = "Distribuicao dos Funcionarios por Grau de Instrucao")


#### Grafico de Pareto ####
if(!require(qcc)) install.packages("qcc", dep = T)
library(qcc)


pareto.chart(table(EsT_Civ))
pareto.chart(table(G_Int))


#############################################################################
############################ DADOS QUANTITATIVOS ############################

#### Medidas de posicao e dispersao ####
# (que sao calculadas em variaveis quantitativas)

# Media
mean(Salario) 

# Variancia
var(Salario)

# Desvio padrao
sd(Salario)

# Mediana
median(Salario)

# Quantis
quantile(Salario, 0.9)

# Quartis
quantile(Salario, c(0.25,0.5,0.75))

# Extrai as medidas descritivas das variaveis
summary(dados)


#### Funcao que determina as Medidas de Posicao e dispersao ####
meuResumo <- function(x)
{
    s <- c(length(x), mean(x, na.rm = T), median(x, na.rm = T),
           sd(x, na.rm = T), var(x, na.rm = T), length(which(is.na(x))))
    names(s) <- c("n", "media", "mediana", "desvio Padrao", 
                  "variancia", "NA")
    return(s)
}


meuResumo(Salario)


#### Grafico de dados discretos ####
plot(table(N_filhos), xlab = "Numero de filhos",
     ylab = "Numero de Funcionarios")

hist(N_filhos)


#### Grafico de pontos ####
stripchart(Salario, xlab = "Salario (em s.m.)", pch = 20, method = "stack")
abline(h = 0.98)
points(mean(Salario), 0.93, pch = 17, col = "red", cex = 2)


#### Grafico de Histograma ####
hist(Salario, breaks = c(4,8,12,16,20,24), xlab = "Sal?rio",
     ylab = "Densidade", prob = T, col = "yellow", ylim = c(0, 0.1),
     main = "Dist. de sal?rio dos funcion?rios  da empresa",
     labels = c("27,8%", "33,3%", "22,2%", "13,8%", "2,8%"))


# ou

h = hist(Salario, breaks = c(4,8,12,16,20,24), xlab = "Sal?rio",
         ylab = "N?mero de Funcion?rios", col = "yellow", axes = F, right = F,
         main = "Dist. de sal?rio dos funcion?rios da empresa",
         labels = c("27,8%", "33,3%", "22,2%", "13,8%", "2,8%"))
axis(1, h$breaks); axis(2, h$counts)


#### Tabela de Frequencia ####
TDF = hist(Salario, breaks = c(4,8,12,16,20,24), right = F, plot = F)


# Frequencia absoluta
fabs = TDF$counts 


# Frequencia relativa
fr = fabs/length(Salario) 


# Construindo a tabela manualmente
saida = cbind(fabs, fr)
dimnames(saida) = list(c("4|-8", "8|-12", "12|-16", "16|-20", "20|-24"),
                       c("f", "fr"))
saida


#### Tabela de Frequencia usando a funcao cut ####
fabs = table(cut(Salario, breaks = seq(4,24,4), right = F, 
             labels = c("4|-8", "8|-12", "12|-16", "16|-20", "20|-24")))

fr = fabs/length(Salario)
saida = cbind(fabs, fr)
dimnames(saida) = list(c("4|-8", "8|-12", "12|-16", "16|-20", "20|-24"),
                       c("f", "fr"))
saida


#### Diagrama de Caixas ####
boxplot(Salario, col = "yellow", ylab = "Salario")
boxplot(Salario, notch = T, xlab = "Salario", horizontal = T, col = "green")
boxplot(Salario ~ EsT_Civ, xlab = "Salario", col = c("green", "yellow"))
boxplot(Salario ~ G_Int, xlab = "Grau de Instrucao", ylab = "Salario",
        names = c("1 Grau", "2 Grau", "Superior"),
        main = "Diagrama dos salarios dos func. por grau de instrucao")


#### Analise bivariada (quantitativo) ####
plot(Salario[complete.cases(N_filhos)], N_filhos[complete.cases(N_filhos)], 
     xlab = "Salario", ylab = "N de filhos", pch = 20)

# Correlacao
cor(Salario, N_filhos)

x = c(1, 2, 3, 4); y = c(2, 4, 6, 8.5)

cor(x, y)

plot(x, y)
lines(lowess(x, y), col = "blue")