Market Basket Analysis
Luis Mendoza A00838524
2024-09-10
Instalar paquetes y llamar librerías
#install.packages ("tidyverse") # Paquete global para manipulación y análisis de datos
library (tidyverse)
#install.packages("dplyr") # Para filtrar datos
library (dplyr)
#install.packages("janitor") # Para examinar y limpiar bases de datos
library (janitor)
#install.packages ("Matrix") #Para trabajar con matrices
library (Matrix)
#install.packages ("arules") #Genera reglas de asociación
library (arules)
#install.packages ("arulesViz") #Visualizar reglas de asociación
library (arulesViz)
#install.packages ( "datasets")
library (datasets)
#install.packages ("plyr")
library (plyr)#Importar la base de datos
#file.choose()
df <- read.csv("/Users/luismendoza/Downloads/abarrotes.csv")#Análisis descriptivo
summary(df)## vcClaveTienda DescGiro Codigo.Barras PLU
## Length:200625 Length:200625 Min. :8.347e+05 Min. : 1.00
## Class :character Class :character 1st Qu.:7.501e+12 1st Qu.: 1.00
## Mode :character Mode :character Median :7.501e+12 Median : 1.00
## Mean :5.950e+12 Mean : 2.11
## 3rd Qu.:7.501e+12 3rd Qu.: 1.00
## Max. :1.750e+13 Max. :30.00
## NA's :199188
## Fecha Hora Marca Fabricante
## Length:200625 Length:200625 Length:200625 Length:200625
## Class :character Class :character Class :character Class :character
## Mode :character Mode :character Mode :character Mode :character
##
##
##
##
## Producto Precio Ult.Costo Unidades
## Length:200625 Min. :-147.00 Min. : 0.38 Min. : 0.200
## Class :character 1st Qu.: 11.00 1st Qu.: 8.46 1st Qu.: 1.000
## Mode :character Median : 16.00 Median : 12.31 Median : 1.000
## Mean : 19.42 Mean : 15.31 Mean : 1.262
## 3rd Qu.: 25.00 3rd Qu.: 19.23 3rd Qu.: 1.000
## Max. :1000.00 Max. :769.23 Max. :96.000
##
## F.Ticket NombreDepartamento NombreFamilia NombreCategoria
## Min. : 1 Length:200625 Length:200625 Length:200625
## 1st Qu.: 33964 Class :character Class :character Class :character
## Median :105993 Mode :character Mode :character Mode :character
## Mean :193990
## 3rd Qu.:383005
## Max. :450040
##
## Estado Mts.2 Tipo.ubicación Giro
## Length:200625 Min. :47.0 Length:200625 Length:200625
## Class :character 1st Qu.:53.0 Class :character Class :character
## Mode :character Median :60.0 Mode :character Mode :character
## Mean :56.6
## 3rd Qu.:60.0
## Max. :62.0
##
## Hora.inicio Hora.cierre
## Length:200625 Length:200625
## Class :character Class :character
## Mode :character Mode :character
##
##
##
## str(df)## 'data.frame': 200625 obs. of 22 variables:
## $ vcClaveTienda : chr "MX001" "MX001" "MX001" "MX001" ...
## $ DescGiro : chr "Abarrotes" "Abarrotes" "Abarrotes" "Abarrotes" ...
## $ Codigo.Barras : num 7.5e+12 7.5e+12 7.5e+12 7.5e+12 7.5e+12 ...
## $ PLU : int NA NA NA NA NA NA NA NA NA NA ...
## $ Fecha : chr "19/06/2020" "19/06/2020" "19/06/2020" "19/06/2020" ...
## $ Hora : chr "08:16:21" "08:23:33" "08:24:33" "08:24:33" ...
## $ Marca : chr "NUTRI LECHE" "DAN UP" "BIMBO" "PEPSI" ...
## $ Fabricante : chr "MEXILAC" "DANONE DE MEXICO" "GRUPO BIMBO" "PEPSI-COLA MEXICANA" ...
## $ Producto : chr "Nutri Leche 1 Litro" "DANUP STRAWBERRY P/BEBER 350GR NAL" "Rebanadas Bimbo 2Pz" "Pepsi N.R. 400Ml" ...
## $ Precio : num 16 14 5 8 19.5 16 14 5 8 19.5 ...
## $ Ult.Costo : num 12.3 14 5 8 15 ...
## $ Unidades : num 1 1 1 1 1 1 1 1 1 1 ...
## $ F.Ticket : int 1 2 3 3 4 1 2 3 3 4 ...
## $ NombreDepartamento: chr "Abarrotes" "Abarrotes" "Abarrotes" "Abarrotes" ...
## $ NombreFamilia : chr "Lacteos y Refrigerados" "Lacteos y Refrigerados" "Pan y Tortilla" "Bebidas" ...
## $ NombreCategoria : chr "Leche" "Yogurt" "Pan Dulce Empaquetado" "Refrescos Plástico (N.R.)" ...
## $ Estado : chr "Nuevo León" "Nuevo León" "Nuevo León" "Nuevo León" ...
## $ Mts.2 : int 60 60 60 60 60 60 60 60 60 60 ...
## $ Tipo.ubicación : chr "Esquina" "Esquina" "Esquina" "Esquina" ...
## $ Giro : chr "Abarrotes" "Abarrotes" "Abarrotes" "Abarrotes" ...
## $ Hora.inicio : chr "08:00" "08:00" "08:00" "08:00" ...
## $ Hora.cierre : chr "22:00" "22:00" "22:00" "22:00" ...#count (df, vcClaveTienda)
#count (df, DescGiro)
#count (df, Fecha)
#count (df, Hora)
#count (df, Marca)
#count (df, Fabricante)
#count (df, Producto)
#count (df, NombreDepartamento)
#count (df, NombreFamilia)
#count (df, NombreCategoria)
#count (df, Estado)
#count (df, Giro)
#count (df, Hora.inicio)
#count (df, Hora.cierre)Tablas
# Tabla de tienda y departamento
tabyl (df, vcClaveTienda, NombreDepartamento)## vcClaveTienda Abarrotes Bebes e Infantiles Carnes Farmacia Ferretería Mercería
## MX001 95415 515 1 147 245 28
## MX002 6590 21 0 4 10 0
## MX003 4026 15 0 2 8 0
## MX004 82234 932 0 102 114 16
## MX005 10014 0 0 0 0 0
## Papelería Productos a Eliminar Vinos y Licores
## 35 3 80
## 0 0 4
## 0 0 0
## 32 5 20
## 7 0 0# Tabla de Estado y Hora de Inicio
tabyl (df, Estado, Hora.inicio)## Estado 07:00 08:00 09:00
## Chiapas 4051 0 0
## Jalisco 0 0 6629
## Nuevo León 0 96469 0
## Quintana Roo 0 10021 0
## Sinaloa 83455 0 0Limpieza de datos
Técnica 1. Eliminar los valores irrelevantes
# Eliminar columnas
#df<- subset (df, select = -c(PLU))
#Eliminar renglones
df<- df[df$Precio > 0, ]Técnica 3. Eliminar los valores repetidos
df <- distinct (df)##Técnica 3.Corregir errores tipográficos y similares
df$Unidades <- ceiling (df$Unidades)
summary (df$Unidades)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 1.000 1.000 1.000 1.261 1.000 96.000##Técnica 4.Convertir tipos de datos
# Convertir de caracter a fecha
df$Fecha <- as.Date (df$Fecha, format="%d/%m/%Y")
str(df$Fecha)## Date[1:200473], format: "2020-06-19" "2020-06-19" "2020-06-19" "2020-06-19" "2020-06-19" ...summary (df$Fecha)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## "2020-05-01" "2020-06-06" "2020-07-11" "2020-07-18" "2020-08-29" "2020-11-11"##Técnica 5.Tratar valores faltantes
# Borrar todos los NAs
#df <- na.omit(df). Puede eliminar otros datos, usar con cuidado
#Reemplazar los NAs con CEROS (No tan común)
#df [is.na(df)] <- 0
#Reemplazar los NAs con el promedio (Más usado)
#df [is.na(df)] <- mean (df$altura, na.rm =TRUE)##Técnica 6.Herramientas estadísticas
# Detectar errores en bases de datos, los que están fuera del rango deben ser revisados
boxplot(df$Precio, horizontal = TRUE)
boxplot(df$Unidades, horizontal = TRUE)
#Generar basket
#Ordenar de menor a mayor la columna Ticket
df <- df[order(df$F.Ticket), ]
head (df)## vcClaveTienda DescGiro Codigo.Barras PLU Fecha Hora
## 1 MX001 Abarrotes 7.501021e+12 NA 2020-06-19 08:16:21
## 2 MX001 Abarrotes 7.501032e+12 NA 2020-06-19 08:23:33
## 3 MX001 Abarrotes 7.501000e+12 NA 2020-06-19 08:24:33
## 4 MX001 Abarrotes 7.501031e+12 NA 2020-06-19 08:24:33
## 5 MX001 Abarrotes 7.501026e+12 NA 2020-06-19 08:26:28
## 6 MX001 Abarrotes 7.501025e+12 NA 2020-06-19 08:26:28
## Marca Fabricante
## 1 NUTRI LECHE MEXILAC
## 2 DAN UP DANONE DE MEXICO
## 3 BIMBO GRUPO BIMBO
## 4 PEPSI PEPSI-COLA MEXICANA
## 5 BLANCA NIEVES (DETERGENTE) FABRICA DE JABON LA CORONA
## 6 FLASH ALEN
## Producto Precio Ult.Costo Unidades F.Ticket
## 1 Nutri Leche 1 Litro 16.0 12.31 1 1
## 2 DANUP STRAWBERRY P/BEBER 350GR NAL 14.0 14.00 1 2
## 3 Rebanadas Bimbo 2Pz 5.0 5.00 1 3
## 4 Pepsi N.R. 400Ml 8.0 8.00 1 3
## 5 Detergente Blanca Nieves 500G 19.5 15.00 1 4
## 6 Flash Xtra Brisa Marina 500Ml 9.5 7.31 1 4
## NombreDepartamento NombreFamilia NombreCategoria
## 1 Abarrotes Lacteos y Refrigerados Leche
## 2 Abarrotes Lacteos y Refrigerados Yogurt
## 3 Abarrotes Pan y Tortilla Pan Dulce Empaquetado
## 4 Abarrotes Bebidas Refrescos Plástico (N.R.)
## 5 Abarrotes Limpieza del Hogar Lavandería
## 6 Abarrotes Limpieza del Hogar Limpiadores Líquidos
## Estado Mts.2 Tipo.ubicación Giro Hora.inicio Hora.cierre
## 1 Nuevo León 60 Esquina Abarrotes 08:00 22:00
## 2 Nuevo León 60 Esquina Abarrotes 08:00 22:00
## 3 Nuevo León 60 Esquina Abarrotes 08:00 22:00
## 4 Nuevo León 60 Esquina Abarrotes 08:00 22:00
## 5 Nuevo León 60 Esquina Abarrotes 08:00 22:00
## 6 Nuevo León 60 Esquina Abarrotes 08:00 22:00tail (df)## vcClaveTienda DescGiro Codigo.Barras PLU Fecha Hora
## 107247 MX004 Carnicería 1.024877e+10 NA 2020-10-15 11:51:40
## 167624 MX004 Carnicería 7.501080e+12 NA 2020-10-15 11:51:40
## 149282 MX004 Carnicería 7.501055e+12 NA 2020-10-15 11:54:37
## 168603 MX004 Carnicería 7.501214e+12 NA 2020-10-15 11:56:52
## 161046 MX004 Carnicería 7.501031e+12 NA 2020-10-15 12:01:54
## 112823 MX004 Carnicería 7.500470e+07 NA 2020-10-15 12:02:36
## Marca Fabricante Producto
## 107247 YEMINA HERDEZ PASTA SPAGHETTI YEMINA 200G
## 167624 DEL FUERTE ALIMENTOS DEL FUERTE PURE DE TOMATE DEL FUERTE 345G
## 149282 COCA COLA ZERO COCA COLA COCA COLA ZERO 600ML
## 168603 DIAMANTE EMPACADOS ARROZ DIAMANTE225G
## 161046 PEPSI PEPSI-COLA MEXICANA PEPSI N. R. 500ML
## 112823 COCA COLA COCA COLA COCA COLA RETORNABLE 500ML
## Precio Ult.Costo Unidades F.Ticket NombreDepartamento
## 107247 7 5.38 2 450032 Abarrotes
## 167624 12 9.23 1 450032 Abarrotes
## 149282 15 11.54 2 450034 Abarrotes
## 168603 11 8.46 1 450037 Abarrotes
## 161046 10 7.69 1 450039 Abarrotes
## 112823 10 7.69 8 450040 Abarrotes
## NombreFamilia NombreCategoria Estado Mts.2
## 107247 Sopas y Pastas Fideos, Spaguetti, Tallarines Sinaloa 53
## 167624 Salsas y Sazonadores Salsa para Spaguetti Sinaloa 53
## 149282 Bebidas Refrescos Retornables Sinaloa 53
## 168603 Granos y Semillas Arroz Sinaloa 53
## 161046 Bebidas Refrescos Plástico (N.R.) Sinaloa 53
## 112823 Bebidas Refrescos Retornables Sinaloa 53
## Tipo.ubicación Giro Hora.inicio Hora.cierre
## 107247 Esquina Abarrotes 07:00 23:00
## 167624 Esquina Abarrotes 07:00 23:00
## 149282 Esquina Abarrotes 07:00 23:00
## 168603 Esquina Abarrotes 07:00 23:00
## 161046 Esquina Abarrotes 07:00 23:00
## 112823 Esquina Abarrotes 07:00 23:00#Generar el basket
basket <- ddply (df,c("F.Ticket"), function (df)paste(df$Marca, collapse = ","))
#Eliminar número de Ticket
basket$F.Ticket <- NULL
#Cambiar el título de la columna V1 por Marca
colnames(basket) <- c("Marca")
#Exportar basket
write.csv (basket, "basket.csv", quote = FALSE, row.names = FALSE)#Market basket
#file.choose()
tr<- read.transactions("/Users/luismendoza/basket.csv", format = "basket", sep= ",")
reglas.asociacion <- apriori (tr, parameter= list(supp= 0.001, conf= 0.2, maxlen= 10))## Apriori
##
## Parameter specification:
## confidence minval smax arem aval originalSupport maxtime support minlen
## 0.2 0.1 1 none FALSE TRUE 5 0.001 1
## maxlen target ext
## 10 rules TRUE
##
## Algorithmic control:
## filter tree heap memopt load sort verbose
## 0.1 TRUE TRUE FALSE TRUE 2 TRUE
##
## Absolute minimum support count: 115
##
## set item appearances ...[0 item(s)] done [0.00s].
## set transactions ...[604 item(s), 115031 transaction(s)] done [0.01s].
## sorting and recoding items ... [207 item(s)] done [0.00s].
## creating transaction tree ... done [0.02s].
## checking subsets of size 1 2 3 done [0.00s].
## writing ... [11 rule(s)] done [0.00s].
## creating S4 object ... done [0.01s].summary (reglas.asociacion)## set of 11 rules
##
## rule length distribution (lhs + rhs):sizes
## 2
## 11
##
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 2 2 2 2 2 2
##
## summary of quality measures:
## support confidence coverage lift
## Min. :0.001017 Min. :0.2069 Min. :0.003564 Min. : 1.326
## 1st Qu.:0.001104 1st Qu.:0.2358 1st Qu.:0.004507 1st Qu.: 1.789
## Median :0.001417 Median :0.2442 Median :0.005807 Median : 3.972
## Mean :0.001521 Mean :0.2537 Mean :0.006056 Mean :17.558
## 3rd Qu.:0.001652 3rd Qu.:0.2685 3rd Qu.:0.006894 3rd Qu.:21.808
## Max. :0.002747 Max. :0.3098 Max. :0.010502 Max. :65.862
## count
## Min. :117.0
## 1st Qu.:127.0
## Median :163.0
## Mean :174.9
## 3rd Qu.:190.0
## Max. :316.0
##
## mining info:
## data ntransactions support confidence
## tr 115031 0.001 0.2
## call
## apriori(data = tr, parameter = list(supp = 0.001, conf = 0.2, maxlen = 10))inspect (reglas.asociacion)## lhs rhs support confidence coverage
## [1] {FANTA} => {COCA COLA} 0.001051890 0.2439516 0.004311881
## [2] {SALVO} => {FABULOSO} 0.001104050 0.3097561 0.003564257
## [3] {FABULOSO} => {SALVO} 0.001104050 0.2347505 0.004703080
## [4] {COCA COLA ZERO} => {COCA COLA} 0.001417009 0.2969035 0.004772627
## [5] {SPRITE} => {COCA COLA} 0.001347463 0.2069426 0.006511288
## [6] {PINOL} => {CLORALEX} 0.001017117 0.2368421 0.004294495
## [7] {BLUE HOUSE} => {BIMBO} 0.001712582 0.2720994 0.006293956
## [8] {HELLMANN´S} => {BIMBO} 0.001538716 0.2649701 0.005807130
## [9] {REYMA} => {CONVERMEX} 0.002095087 0.2441743 0.008580296
## [10] {FUD} => {BIMBO} 0.001590876 0.2186380 0.007276299
## [11] {COCA COLA LIGHT} => {COCA COLA} 0.002747086 0.2615894 0.010501517
## lift count
## [1] 1.562646 121
## [2] 65.862391 127
## [3] 65.862391 127
## [4] 1.901832 163
## [5] 1.325583 155
## [6] 25.063647 117
## [7] 4.078691 197
## [8] 3.971823 177
## [9] 18.551922 241
## [10] 3.277319 183
## [11] 1.675626 316reglas.asociacion <- sort(reglas.asociacion, by= "confidence", dicreasing =TRUE )
summary (reglas.asociacion)## set of 11 rules
##
## rule length distribution (lhs + rhs):sizes
## 2
## 11
##
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 2 2 2 2 2 2
##
## summary of quality measures:
## support confidence coverage lift
## Min. :0.001017 Min. :0.2069 Min. :0.003564 Min. : 1.326
## 1st Qu.:0.001104 1st Qu.:0.2358 1st Qu.:0.004507 1st Qu.: 1.789
## Median :0.001417 Median :0.2442 Median :0.005807 Median : 3.972
## Mean :0.001521 Mean :0.2537 Mean :0.006056 Mean :17.558
## 3rd Qu.:0.001652 3rd Qu.:0.2685 3rd Qu.:0.006894 3rd Qu.:21.808
## Max. :0.002747 Max. :0.3098 Max. :0.010502 Max. :65.862
## count
## Min. :117.0
## 1st Qu.:127.0
## Median :163.0
## Mean :174.9
## 3rd Qu.:190.0
## Max. :316.0
##
## mining info:
## data ntransactions support confidence
## tr 115031 0.001 0.2
## call
## apriori(data = tr, parameter = list(supp = 0.001, conf = 0.2, maxlen = 10))inspect (reglas.asociacion)## lhs rhs support confidence coverage
## [1] {SALVO} => {FABULOSO} 0.001104050 0.3097561 0.003564257
## [2] {COCA COLA ZERO} => {COCA COLA} 0.001417009 0.2969035 0.004772627
## [3] {BLUE HOUSE} => {BIMBO} 0.001712582 0.2720994 0.006293956
## [4] {HELLMANN´S} => {BIMBO} 0.001538716 0.2649701 0.005807130
## [5] {COCA COLA LIGHT} => {COCA COLA} 0.002747086 0.2615894 0.010501517
## [6] {REYMA} => {CONVERMEX} 0.002095087 0.2441743 0.008580296
## [7] {FANTA} => {COCA COLA} 0.001051890 0.2439516 0.004311881
## [8] {PINOL} => {CLORALEX} 0.001017117 0.2368421 0.004294495
## [9] {FABULOSO} => {SALVO} 0.001104050 0.2347505 0.004703080
## [10] {FUD} => {BIMBO} 0.001590876 0.2186380 0.007276299
## [11] {SPRITE} => {COCA COLA} 0.001347463 0.2069426 0.006511288
## lift count
## [1] 65.862391 127
## [2] 1.901832 163
## [3] 4.078691 197
## [4] 3.971823 177
## [5] 1.675626 316
## [6] 18.551922 241
## [7] 1.562646 121
## [8] 25.063647 117
## [9] 65.862391 127
## [10] 3.277319 183
## [11] 1.325583 155top10reglas <- head (reglas.asociacion, n=10, by= "confidence")
plot(top10reglas, method= "graph", engine = "htmlwidget")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