library(DT)
tablet=tibble(turno=gl(2,1,2,c("diurno","noche")), items= c(11000,4000),
          prob_defect= c(0.04,0.08),prob_exi= c(1-c(0.04,0.08)),
          "peso(%)"= round(c(11000,4000)/sum(11000,4000),2)*100,
          D= round((0.10^2)/(1.96^2),4),tiempo=c(480,480))
datatable(tablet)

\[n={\sum_{i=1}^3 N_{i}^2 p_{i}(1-p_{i})/w_{i}\over{\sum_{i=1}^3 N_{i}p_{i}(1-p_{i})}+({Ne\over{z}})^2} \]

# utilizando for 
muestreo_estratificado<-function(N,p,z=1.96,e=0.05,t){
  Ni=sum(N)
  w <<-N/Ni
   numerador<<-c()
    for(i in 1:length(N)){
    numerador[i]<-(N[i]^2*p[i]*(1-p[i])/w[i])
    }
  denominador=c()
    for(i in 1:length(N)){
    denominador[i]<-(N[i]*p[i]*(1-p[i]))
    }
  n=ceiling(sum(numerador)/(sum(denominador)+(Ni*e/z)^2))
  distri<-round(w*n)
  k = floor(t / distri) # Tiempo de muestreo sistematico por estrato
  return(list(n,distri,k))
}
mm=muestreo_estratificado(N=tablet$items,p=tablet$prob_defect,t=tablet$tiempo);mm
## [[1]]
## [1] 74
## 
## [[2]]
## [1] 54 20
## 
## [[3]]
## [1]  8 24
library(DT)

pb=c(((11/5)*0.03)+((4/15)*0.06))

tablet2=tibble(Turno=gl(2,1,2,c("diurno","noche")), 
               "items(und)"= c(11000,4000),
          "prob_defect_turno(%)"=round(c(((11/5)*0.03)/pb,
                             ((4/15)*0.06)/pb)*100,2),
          "Muestras(und)"= mm[[2]] ,"tiempo(Hr)"=mm[[3]])
datatable(tablet2)

Parte 2. Muestreo Estratificado Espacial (Hipercubo Latino Condicionado)

Ejercicio. Datos proveneientes de cultivo de palma

set.seed(123)
palmas <- expand.grid(x=seq(0,112,7), y=seq(0,144,9))
nrow(palmas)
## [1] 289
## Peso promedio de los racimos ultima cosecha
p_racimo_u <- rnorm(289,17,1.8)
## Peso promedio de los racimos penultima cosecha
p_racimo_pe <- rnorm(289,17,1.8)
## Relación Ca/Mg Hoja 17
CaMg_h17<-runif(289,1.8,2)
## Relación Ca/Mg Suelo
CaMg_s<-runif(289,1.2,1.4)
## Hibridos (ejemplo de una variable cualitativa)
hibrid <- rep(c('h1','h2'),c(144,145))

# 4 variables auxiliares equivalentes a 4 dimensiones (Hipercubo)

# Articulo base. A conditionated Latin hypercube method 

dfpalma<-data.frame(palmas,p_racimo_u,p_racimo_pe,CaMg_h17,CaMg_s,hibrid)
datatable(dfpalma)
set.seed(2022)
library(clhs)
res<-clhs(dfpalma,size=25,progress = FALSE,simple = TRUE)
## Warning: NAs introduced by coercion
res2= ifelse(res>=144,"H2","H1")
dfpalma[res,'muestreo']= res2
dfpalma$muestreo[is.na(dfpalma$muestreo)]='no'
library(ggplot2)
ggplot(dfpalma)+aes(x,y,fill=muestreo)+geom_tile(color='white')

library(dplyr)
dfpalma %>% 
  group_by(hibrid) %>% 
  summarise(mean_PR= mean(p_racimo_pe),
            mean_u= mean(p_racimo_u),
            mean_CaMg_h17 = mean(CaMg_h17),
            mean_CaMg_s = mean(dfpalma$CaMg_s)) %>% 
  data.frame()-> hibrido

datatable(hibrido)
library(dplyr)
dfpalma %>% 
  group_by(muestreo) %>% 
  summarise(mean_PR= mean(p_racimo_pe),
            mean_u= mean(p_racimo_u),
            mean_CaMg_h17 = mean(CaMg_h17),
            mean_CaMg_s = mean(dfpalma$CaMg_s)) %>% 
  data.frame()-> hibrido2

datatable(hibrido2)