Analisis de Valores Core Sampler

Se realizó un análisis horario de los valores de Pol, Brix y Pureza reportados por el Core Sampler.

El objetivo es verificar la hipótesis de que durante cierto rango de horas del día los valores de Pol, Brix y Pureza tienden a ser más altos que en el resto de horas. Específicamente, durante las primeras horas del día.

library(ggplot2)
library(lubridate)
library(dplyr)

Conjunto de Datos Muestras Core Sampler

Este conjunto de datos se obtuvo de la base de datos Legacy de Laboratorio

dataset <- read.csv(file = 'C:/Users/100346/OneDrive - Pantaleon. S.A/DataScience Pantaleon/R/Core Sampler/Datos_Core_Z2025.csv')
dataset$Fecha <- as.Date(dataset$Fecha,format = "%d/%m/%Y")
dataset

Verificación de Rangos

Se examinan los promedios de Pol, Brix y Pureza para los siguientes rangos horarios:

• Rango 1: 0 a 7 horas

• Rango 2: 8 a 15 horas

• Rango 3: 16 a 23 horas Y, para cada Rango, se analizan distintos horizontes temporales:

• Horizonte 1: 23 NOV al 30 NOV

• Horizonte 2: 23 NOV al 15 DIC

• Horizonte 3: 23 NOV al 16 ENE

df1 <- dataset %>% filter(Fecha < '2024-11-30')
df1_dist <- df1 %>%  group_by(hora) %>% summarise(Pol = mean(Pol, na.rm = TRUE), Brix = mean(Brix, na.rm = TRUE), 
                                      Pureza = mean(Pureza, na.rm = TRUE) , ensayos = n())

df2 <- dataset %>% filter(Fecha < '2024-12-15')
df2_dist <- df2 %>%  group_by(hora) %>% summarise(Pol = mean(Pol, na.rm = TRUE), Brix = mean(Brix, na.rm = TRUE), 
                                      Pureza = mean(Pureza, na.rm = TRUE) , ensayos = n())

df3 <- dataset %>% filter(Fecha < '2025-01-16')
df3_dist <- df3 %>%  group_by(hora) %>% summarise(Pol = mean(Pol, na.rm = TRUE), Brix = mean(Brix, na.rm = TRUE), 
                                      Pureza = mean(Pureza, na.rm = TRUE) , ensayos = n())

Promedios Horizonte 1: 23 NOV al 30 NOV

NOV30_Primero <- df1_dist %>% filter(hora %in% (0:7) ) %>% 
  summarise(Pol_Promedio = mean(Pol, na.rm = TRUE), Brix_Promedio = mean(Brix, na.rm = TRUE), Pureza_Promedio = mean(Pureza, na.rm = TRUE))

NOV30_Segundo <- df1_dist %>% filter(hora %in% (8:15) ) %>% 
  summarise(Pol_Promedio = mean(Pol, na.rm = TRUE), Brix_Promedio = mean(Brix, na.rm = TRUE), Pureza_Promedio = mean(Pureza, na.rm = TRUE))

NOV30_Tercero <- df1_dist %>% filter(hora %in% (16:23) ) %>% 
  summarise(Pol_Promedio = mean(Pol, na.rm = TRUE), Brix_Promedio = mean(Brix, na.rm = TRUE), Pureza_Promedio = mean(Pureza, na.rm = TRUE))

df1_Pol <- data.frame("Variable" = c("Pol","Brix","Pureza"), 
                      "de 0 a 7 horas" = c(NOV30_Primero$Pol_Promedio,NOV30_Primero$Brix_Promedio, NOV30_Primero$Pureza_Promedio),
                      "de 8 a 15 horas" = c(NOV30_Segundo$Pol_Promedio,NOV30_Segundo$Brix_Promedio, NOV30_Segundo$Pureza_Promedio),
                      "de 16 a 23 horas" = c(NOV30_Tercero$Pol_Promedio,NOV30_Tercero$Brix_Promedio, NOV30_Tercero$Pureza_Promedio)
                      )
df1_Pol

Promedios Horizonte 2: 23 NOV al 14 DIC

DIC15_Primero <- df2_dist %>% filter(hora %in% (0:7) ) %>% 
  summarise(Pol_Promedio = mean(Pol, na.rm = TRUE), Brix_Promedio = mean(Brix, na.rm = TRUE), Pureza_Promedio = mean(Pureza, na.rm = TRUE))

DIC15_Segundo <- df2_dist %>% filter(hora %in% (8:15) ) %>% 
  summarise(Pol_Promedio = mean(Pol, na.rm = TRUE), Brix_Promedio = mean(Brix, na.rm = TRUE), Pureza_Promedio = mean(Pureza, na.rm = TRUE))

DIC15_Tercero <- df2_dist %>% filter(hora %in% (16:23) ) %>% 
  summarise(Pol_Promedio = mean(Pol, na.rm = TRUE), Brix_Promedio = mean(Brix, na.rm = TRUE), Pureza_Promedio = mean(Pureza, na.rm = TRUE))

df2_Pol <- data.frame("Variable" = c("Pol","Brix","Pureza"), 
                      "de 0 a 7 horas" = c(DIC15_Primero$Pol_Promedio,DIC15_Primero$Brix_Promedio, DIC15_Primero$Pureza_Promedio),
                      "de 8 a 15 horas" = c(DIC15_Segundo$Pol_Promedio,DIC15_Segundo$Brix_Promedio, DIC15_Segundo$Pureza_Promedio),
                      "de 16 a 23 horas" = c(DIC15_Tercero$Pol_Promedio,DIC15_Tercero$Brix_Promedio, DIC15_Tercero$Pureza_Promedio)
                      )
df2_Pol

Promedios Horizonte 3: 23 NOV al 16 ENE

ENE16_Primero <- df3_dist %>% filter(hora %in% (0:7) ) %>% 
  summarise(Pol_Promedio = mean(Pol, na.rm = TRUE), Brix_Promedio = mean(Brix, na.rm = TRUE), Pureza_Promedio = mean(Pureza, na.rm = TRUE))

ENE16_Segundo <- df3_dist %>% filter(hora %in% (8:15) ) %>% 
  summarise(Pol_Promedio = mean(Pol, na.rm = TRUE), Brix_Promedio = mean(Brix, na.rm = TRUE), Pureza_Promedio = mean(Pureza, na.rm = TRUE))

ENE16_Tercero <- df3_dist %>% filter(hora %in% (16:23) ) %>% 
  summarise(Pol_Promedio = mean(Pol, na.rm = TRUE), Brix_Promedio = mean(Brix, na.rm = TRUE), Pureza_Promedio = mean(Pureza, na.rm = TRUE))

df3_Pol <- data.frame("Variable" = c("Pol","Brix","Pureza"), 
                      "de 0 a 7 horas" = c(ENE16_Primero$Pol_Promedio,ENE16_Primero$Brix_Promedio, ENE16_Primero$Pureza_Promedio),
                      "de 8 a 15 horas" = c(ENE16_Segundo$Pol_Promedio,ENE16_Segundo$Brix_Promedio, ENE16_Segundo$Pureza_Promedio),
                      "de 16 a 23 horas" = c(ENE16_Tercero$Pol_Promedio,ENE16_Tercero$Brix_Promedio, ENE16_Tercero$Pureza_Promedio)
                      )
df3_Pol

Número de muestreos por rango de horas

# Bar Plots:
barplot(height=df3_dist$ensayos, names=df3_dist$hora, col = "light green", xlab = "Hora Muestreo", ylab = "Acumulado Muestras", cex.axis=0.5, cex.names=0.5)
barplot(height=df2_dist$ensayos, names=df2_dist$hora, col = "green", cex.axis=0.5, cex.names=0.5, add = TRUE)

barplot(height=df1_dist$ensayos, names=df1_dist$hora, col = "dark green", cex.axis=0.5, cex.names=0.5, add = TRUE)

legend("top", title="Período", legend= c("al 30-Nov","al 15-Dic","al 16-Ene"), fill =c("dark green", "green", "light green" ), box.lty=0)

df_dist <- dataset %>%  group_by(hora) %>% summarise(Pol = mean(Pol, na.rm = TRUE), Brix = mean(Brix, na.rm = TRUE), 
                                      Pureza = mean(Pureza, na.rm = TRUE) , Rendimiento = mean(Rendimiento, na.rm = TRUE), ensayos = n())

# Bar Plots:
barplot(height=df_dist$Pol, names=df_dist$hora, col = "darkorange", xlab = "Hora Muestreo", ylab = "Pol", cex.axis=0.5, cex.names=0.5)

title(main="Distribución Horaria de Pol")

# Bar Plots:
barplot(height=df_dist$Brix, names=df_dist$hora, col = "darkgreen", xlab = "Hora Muestreo", ylab = "Brix", cex.axis=0.5, cex.names=0.5)
title(main="Distribución Horaria de Brix")

# Bar Plots:
barplot(height=df_dist$Pureza, names=df_dist$hora, col = "blue", xlab = "Hora Muestreo", ylab = "Pureza", cex.axis=0.5, cex.names=0.5)
title(main="Distribución Horaria de Pureza")

# Bar Plots:
barplot(height=df_dist$Rendimiento, names=df_dist$hora, col = "darkred", xlab = "Hora Muestreo", ylab = "Rendimiento", cex.axis=0.5, cex.names=0.5)
title(main="Distribución Horaria de Rendimiento Core")

• No existe evidencia de que la Pol, Brix, Pureza o Rendimiento del Core Sampler tenga diferencia significativa durante el día.
• No existe evidencia que la distribución de registro de muestras a lo largo del día esté sesgada.