Clase6

#Variable cuantitativa y cualitativa

##Variables caulitativas #categoricas y se dividen en dos: ordinales y nominales

#Variables cuantitativas #expresar numericamente, discretas y continuas

library(dplyr)

## 
## Adjuntando el paquete: 'dplyr'

## The following objects are masked from 'package:stats':
## 
##     filter, lag

## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union

library(dslabs)
library(viridisLite)

#Dataset

data("murders")

#Graficos

##Diagrama de barras

levels(murders$region)

## [1] "Northeast"     "South"         "North Central" "West"

tabla <- table(murders$region)

barplot(tabla, main = "Grafico de barras", ylab = "Frecuencia", xlab = "Regiones", ylim = c(0,20),col = "red", ) %>% text(y = tabla, label = tabla, pos = 3, cex = 0.8, col = "blue")

help("barplot")

## starting httpd help server ... done

#Diagrama de torta

pie(tabla, labels = row.names(tabla), col = viridis(4), main = "Diagrama circular", borde = viridis(4))
#col = el nombre de la paleta y el numero
porcentajes <- round(tabla * 100/ sum(tabla) , 2)
etiquetas <- paste0(murders$region, " (", porcentajes, ")")
legend("topright",legend=names(tabla),cex=0.75,fill=viridis(4))

#crear tabla de porcentajes y graficar mostrando el dato

requiero un orden especifico

region<-factor(murders$region, levels=c("South", "West", "North Central", "Northeast"))
tabla1<-table(region)
barplot(tabla1)

#Histograma

murders <- murders  %>% mutate(murders, rate = total/population*100000)

hist(murders$rate, main = "Histograma tasa de asesinatos", ylab = "Frecuencia", xlab = "Tasa de asesinatos", xlim = c(0,20), ylim = c(0,25), col = rocket(4)) 

#Grafico grafico de cajas

boxplot(murders$rate, col="Blue",ylab="Tasa de asesinatos",outline=FALSE,main="Boxplot",ylim=c(0,6))

#Adicionar la media

points(mean(murders$rate),col="black",pch=20)
text(paste(" ", round(mean(murders$rate), 2)),x=1.1,y=2.3) 

#comparacion variables cuantitativas y cualitativas

#diagrama de dispersion total vs poblacion

x<-murders$population/10^6
y<-murders$total
plot(x,y,main="Total vs Population",col=murders$region,pch=20,xlab = "Population/10^6",ylab="Population")
legend("bottomright",legend=levels(murders$region),fill = unique(murders$region))

#bolxplot rate vs region
boxplot(rate~region,data=murders,col=rocket(4),outline=FALSE,ylim=c(0,6))
abline(h=mean(murders$rate),col="red",lwd=2)
text(paste("",round(mean(murders$rate),2)),x=0.5,y=3,col="green")

#varios graficos

par(mfrow=c(1,2))

plot(x,y,main="Total vs Population",col=murders$region,pch=20,xlab = "Population/10^6",ylab="Population")
legend("bottomright",legend=levels(murders$region),fill = unique(murders$region))

boxplot(rate~region,data=murders,col=rocket(4),outline=FALSE,ylim=c(0,6))
abline(h=mean(murders$rate),col="red",lwd=2)
text(paste("",round(mean(murders$rate),2)),x=0.7,y=3,col="green")