A-INTRODUCCION.
Estos datos representan la tabla por volumen de ventas y precio promedio del aguacate hass en USA en el periodo entre 2015-2020, tanto a nivel general (“Total US”), como los datos por regiones en USA que incluyen 8 grandes regiones (California, West, South Central, Great Lakes, Midsouth, Southeast,Northeast, Plains), como a nivel de mercado en ciudades USA.
B-OBJETO DEL ESTUDIO.
Analisis exploratorio y conclusiones del mismo, asi como la prediccion Machine Learning para estimar un modelo que prediga el precio promedio del aguacate y por ultimo hacer prediciones tanto de precio promedio como del volumen de ventas del aguacate por tipos (organico y convencional) a traves del metodo ARIMA de Series temporales en un horizonte a futuro de 96 semanas.
1. ANALISIS EXPLORATORIO.
library(tidyverse)
library(tseries)
library(lubridate)
library(scales)
library(zoo)
library(caret)
library(forecast)avoc <- read.csv("avocado-2020.csv")str(avoc)## 'data.frame': 30021 obs. of 13 variables:
## $ date : chr "2015-01-04" "2015-01-04" "2015-01-04" "2015-01-04" ...
## $ average_price: num 1.22 1.79 1 1.76 1.08 1.29 1.01 1.64 1.02 1.83 ...
## $ total_volume : num 40873 1374 435021 3847 788025 ...
## $ X4046 : num 2819.5 57.4 364302.4 1500.2 53987.3 ...
## $ X4225 : num 28287 154 23821 938 552906 ...
## $ X4770 : num 49.9 0 82.2 0 39995 ...
## $ total_bags : num 9716 1163 46816 1408 141137 ...
## $ small_bags : num 9187 1163 16707 1071 137146 ...
## $ large_bags : num 530 0 30109 337 3991 ...
## $ xlarge_bags : num 0 0 0 0 0 0 0 0 0 0 ...
## $ type : chr "conventional" "organic" "conventional" "organic" ...
## $ year : int 2015 2015 2015 2015 2015 2015 2015 2015 2015 2015 ...
## $ geography : chr "Albany" "Albany" "Atlanta" "Atlanta" ...summary(avoc)## date average_price total_volume X4046
## Length:30021 Min. :0.440 Min. : 85 Min. : 0
## Class :character 1st Qu.:1.110 1st Qu.: 14299 1st Qu.: 783
## Mode :character Median :1.350 Median : 124205 Median : 10523
## Mean :1.391 Mean : 939255 Mean : 299107
## 3rd Qu.:1.630 3rd Qu.: 489803 3rd Qu.: 115156
## Max. :3.250 Max. :63716144 Max. :22743616
## X4225 X4770 total_bags small_bags
## Min. : 0 Min. : 0.0 Min. : 0 Min. : 0
## 1st Qu.: 2814 1st Qu.: 0.0 1st Qu.: 8374 1st Qu.: 5956
## Median : 24567 Median : 186.8 Median : 50391 Median : 34255
## Mean : 284901 Mean : 21629.4 Mean : 333534 Mean : 232126
## 3rd Qu.: 140947 3rd Qu.: 5424.2 3rd Qu.: 159174 3rd Qu.: 112938
## Max. :20470573 Max. :2546439.1 Max. :31689189 Max. :20550407
## large_bags xlarge_bags type year
## Min. : 0 Min. : 0 Length:30021 Min. :2015
## 1st Qu.: 352 1st Qu.: 0 Class :character 1st Qu.:2016
## Median : 5171 Median : 0 Mode :character Median :2017
## Mean : 95185 Mean : 6223 Mean :2017
## 3rd Qu.: 36068 3rd Qu.: 560 3rd Qu.:2019
## Max. :13327601 Max. :1022565 Max. :2020
## geography
## Length:30021
## Class :character
## Mode :character
##
##
## Pasamos la variable Date a tipo fecha que utilizaremos a posteriori para nuestro analisis.
avoc$date <- as.Date(avoc$date)Comprobamos si hay elementos NA en nuestro dataset:
colSums(is.na(avoc))## date average_price total_volume X4046 X4225
## 0 0 0 0 0
## X4770 total_bags small_bags large_bags xlarge_bags
## 0 0 0 0 0
## type year geography
## 0 0 0y vemos que no existen valores NA y en su estructura no hay valores extraños.
Pasaremos en nuestro analisis a hacer un desglose entre los datos por region,
ciudades, y total US que se encuentran en nuestros datos como 3 bloques difentes para orientar nuestro analisis.
Datos por Region.
avoc_region <- avoc %>%
filter(geography %in% c("California", "West", "South Central", "Great Lakes",
"Midsouth", "Southeast", "Northeast", "Plains"))Datos por Ciudad.
avoc_city <- avoc %>%
filter(!(geography %in% c("California", "West", "South Central", "Great Lakes",
"Midsouth", "Southeast", "Northeast", "Plains",
"Total U.S.")))Datos Total US.
avoc_total_us <- avoc%>%filter(geography=="Total U.S.")Vamos ahora a agrupar los datos totales “Total U.S”, para calcular el precio promedio del aguacate del año y mes respectivos en funcion del tipo de los mismos.
avocado2 <- avoc_total_us %>%
mutate_at(vars(date), list(month=month))%>%
group_by(year,month, type) %>%
summarise(totalaverage = mean(`average_price`))Una vez hecho esto, vamos a proceder a unir año y fecha del dataframe anterior en una sola variable, para simplificar y adecuar los datos.
avocado3 <- unite(avocado2,Fecha,c(1:2), sep="-")Aqui vamos a convertir esta variable con la fecha en formato character al formato fecha.
avocado3$Periodo <- as.yearmon(as.character(avocado3$Fecha), "%Y-%m")Una vez ya tenemos la nueva variable Periodo con nuestra fecha en su formato adecuado ya podemos eliminar la otra variable fecha que ya no nos sirve.
avocado3$Fecha=NULLAhora vamos a representar en nuestra grafica el precio promedio del aguacate por año, mes y tipo a nivel general.
ggplot(avocado3,aes(factor(Periodo),totalaverage,group=type, col = type))+
geom_line()+theme(axis.text.x = element_text(angle = 90, hjust = 1))+labs(x=
'year/month', y='AveragePrice')+ggtitle('AVERAGEPRICE AVOCADOS TOTAL US BY TYPE 2015-2020')
Para ver las diferencias de precios del “Total US” entre ambos tipos organico y convencional, realizaremos una grafica con boxplot en la que compararemos las medias de sus promedios de precio respectivamente.
ggplot(avoc_total_us, aes(x = type,y = average_price, fill=type)) +
geom_boxplot()+scale_y_continuous(limits=c(0,2.5),breaks = seq(0.0, 2.5, by=0.2))+labs(x="Type Avocados", y ="Averageprice")+
ggtitle("AVERAGE PRICE DIFFERENCES BETWEEN CONVENTIONAL AND ORGANIC AVOCADO")+
theme_minimal()
Veremos tambien la comparacion entre regiones y ciudades en cuanto al precio promedio y al tipo de avocado con sus correspondientes graficas.
avocado_region_pr <-avoc_region %>%
group_by(year,type,geography) %>%
summarise(totalaverage = mean(`average_price`))%>%
group_by(geography,type)%>%
summarise(totalavg = mean(`totalaverage`))avocado_city_pr <-avoc_city %>%
group_by(year,type,geography) %>%
summarise(totalaverage = mean(`average_price`))%>%
group_by(geography,type)%>%
summarise(totalavg = mean(`totalaverage`))ggplot(avocado_region_pr,aes(geography,totalavg,fill=type))+
geom_bar(position = "dodge",
stat="identity") + scale_y_continuous(labels=scales::comma)+
labs(x='Regions', y='Averageprice')+
ggtitle('AVERAGE PRICE USA AVOCADO BY REGION AND TYPE 2015-2020') 
ggplot(avocado_city_pr,aes(geography,totalavg,fill=type))+
geom_bar(width=0.8, position =position_dodge(0.5),
stat="identity") + scale_y_continuous(labels=scales::comma)+coord_flip()+
labs(x='Regions', y='Averageprice')+
ggtitle('AVERAGE PRICE USA AVOCADO BY CITY AND TYPE 2015-2020') 
Ahora analizaremos los datos del volumen de ventas anual del aguacate por año y tipo considerado.
avocado5 <- avoc_total_us %>%
group_by(year,type) %>%
summarise(totalVolume = sum(`total_volume`)) %>%
mutate(Percentage = totalVolume/sum(totalVolume),
percentLabel = scales::percent(Percentage,
accuracy = 0.01,
decimal.mark = "."))Vamos a reprentar a posteriori graficamente el volumen de ventas del aguacate por tipo y por año a nivel general, desglosando tambien en porcentaje los mismos.
ggplot(data = avocado5,
aes(x = year,
y = totalVolume,
fill = type))+
geom_bar(position = "dodge2",
stat="identity") + coord_flip()+scale_y_continuous(labels=scales::comma)+
geom_text(aes(label = percentLabel),
size = 3,
position = position_dodge2(width = 1))+labs(x='years period', y='sales volume')+
ggtitle('ANNUAL VOLUME OF SALES AVOCADO USA 2015-2020')
Vamos a emplear el desglose de nuestros datos para visualizar por region y ciudad el volumen de ventas de aguacates segun su tamaño.
avocado_region_vol <-avoc_region %>%
group_by(geography) %>%
mutate(rest=total_volume-X4046-X4225-X4770) %>%
summarise(Hass_small = sum(`X4046`), Hass_large= sum(`X4225`), Hass_extralarge= sum(`X4770`),rest= sum(`rest`)) %>%
gather(type_size,volume_size,c(Hass_small, Hass_large, Hass_extralarge,rest)) %>%
group_by(geography,type_size) %>%
summarise(volume_t=sum(`volume_size`)) %>%
mutate(Percentage = volume_t/sum(volume_t),
percentLabel = scales::percent(Percentage,
accuracy = 0.01,
decimal.mark = "."))avocado_city_vol <-avoc_city %>%
group_by(geography) %>%
mutate(rest=total_volume-X4046-X4225-X4770) %>%
summarise(Hass_small = sum(`X4046`), Hass_large= sum(`X4225`), Hass_extralarge= sum(`X4770`),rest= sum(`rest`)) %>%
gather(type_size,volume_size,c(Hass_small, Hass_large, Hass_extralarge,rest)) %>%
group_by(geography,type_size) %>%
summarise(volume_t=sum(`volume_size`)) %>%
mutate(Percentage = volume_t/sum(volume_t),
percentLabel = scales::percent(Percentage,
accuracy = 0.01,
decimal.mark = "."))Representamos graficamente el desglose por region y volumen de ventas de los agua cates de los diferentes tamaños a considerar.
ggplot(data = avocado_region_vol,
aes(reorder(geography,-volume_t),volume_t,fill=type_size))+
geom_bar(position = "stack",
stat="identity") + scale_y_continuous(labels=scales::comma)+
geom_text(aes(label = percentLabel),
size = 2.5,
position =position_stack(vjust = 0.7))+labs(x='Regions', y='Sales Volume')+
ggtitle('US AVOCADO SALES VOLUME BY REGION AND SIZE 2015-2020') 
representamos ese mismo desglose pero ahora por ciudad.
ggplot(data = avocado_city_vol,
aes(reorder(geography,volume_t),volume_t,fill=type_size))+
geom_bar(position = "stack",
stat="identity") +coord_flip()+ scale_y_continuous(labels=scales::comma)+
labs(x='Cities', y='Sales Volume')+
ggtitle('US AVOCADO SALES VOLUME BY CITY AND SIZE 2015-2020')
CONCLUSIONES
En lineas generales el volumen de ventas USA, en funcion del tipo de aguacate (Organico y Convencional), es muy superior los del tipo convencional al organico, aproximandamente cercano al 4% anual el organico con respecto al convencional que es de aproximadamente un 96%.
Se puede concluir que la diferencia entre el precio promedio del aguacate organico en USA con respecto al convencional es de aproximadamente 50 centavos mas superior el organico al convencional.
El precio por regiones USA es superior sobre el resto en las regiones de California y Northeast, sobre todo el precio organico llegando al maximo entre 1.75 y 1.80 dolares de promedio.
El precio promedio por ciudad es destacable en ciudades como San Francisco y Hartford/Springfield sobre todo en el precio del aguacate tipo organico superando los 2 dolares ampliamente, tambien siendo destacado junto con otras en cuanto al precio del tipo convencional estando aproximadamente sobre 1,40 dolares.
En cuanto al volumen de ventas anual de aguacate en USA los años 2018 y 2019 fueron los mas prolificos estando en 2018 sobre los 2000 millones en ventas y en el año 2019 superandolos con suficiencia.
En cuanto al volumen de ventas de aguacates por Regiones en USA, destacan por encima del resto, las regiones West, South Central y California que sobrepasan las 1750 millones de unidades vendidas.
En cuanto a volumen de ventas en ciudades USA, Los Angeles, Nueva York, Dallas/FtWorth, Houston y Phoenix/Tucson son las mas destacadas Estando la ciudad de los Angeles muy por encima del resto cercano a los 900 millones en ventas, en lineas generales el Hass pequeño es el que mas se suele vender en el mercado de las ciudades.
2. MODELO ARIMA PARA PREDICCION PRECIO PROMEDIO y VOLUMEN VENTAS POR TIPOS DE AGUACATE.
Ahora vamos a realizar la prediccion con el modelo arima para los tipos de aguacates de nuestros datos organic y conventional:
avoc_arima <- avoc %>% select(date,average_price,total_volume,type)organic <- avoc_arima %>%
filter(type == "organic") %>%
group_by(date) %>%
summarise(avgpr_org= mean(average_price), avgvol_org=mean(total_volume))conventional <- avoc_arima %>%
filter(type == "conventional") %>%
group_by(date) %>%
summarise(avgpr_conv= mean(average_price), avgvol_conv=mean(total_volume))Realizamos nuestras series temporales con los datos arriba mencionados:
organic_avgpr <- ts(organic$avgpr_org, start=c(2015,1), frequency=52)
organic_avgvol <- ts(organic$avgvol_org,start=c(2015,1), frequency=52)
conventional_avgpr <- ts(conventional$avgpr_conv, start=c(2015,1), frequency=52)
conventional_avgvol <- ts(conventional$avgvol_conv, start=c(2015,1), frequency=52)y realizaremos las predicciones para el mejor modelo arima para cada caso en este caso lo haremos para 96 semanas a futuro con un nivel de confianza del 95%.
2.1 PREDICCION PRECIO PROMEDIO PARA AGUACATE TIPO ORGANICO
arima_avg_org <- auto.arima(organic_avgpr,d=1,D=1, trace=T, stepwise=F, approximation=F)forecast_mod <- forecast(arima_avg_org, h=96,level=95)Plot de la prediccion
plot(forecast_mod)
plot con el ajuste de la prediccion.
autoplot(forecast_mod)+autolayer(fitted(forecast_mod), series="Ajuste")
Sumario resultados
print(summary(forecast_mod))##
## Forecast method: ARIMA(1,1,3)(1,1,0)[52]
##
## Model Information:
## Series: organic_avgpr
## ARIMA(1,1,3)(1,1,0)[52]
##
## Coefficients:
## ar1 ma1 ma2 ma3 sar1
## -0.7566 0.8566 -0.1082 -0.3570 -0.5424
## s.e. 0.0895 0.1023 0.0866 0.0669 0.0588
##
## sigma^2 = 0.004938: log likelihood = 271.37
## AIC=-530.75 AICc=-530.36 BIC=-510.25
##
## Error measures:
## ME RMSE MAE MPE MAPE MASE
## Training set 0.0003346908 0.06251377 0.04371864 -0.0210524 2.698541 0.2486772
## ACF1
## Training set -0.01280563
##
## Forecasts:
## Point Forecast Lo 95 Hi 95
## 2020.346 1.577947 1.440214655 1.715679
## 2020.365 1.496456 1.291695903 1.701216
## 2020.385 1.488244 1.247726105 1.728763
## 2020.404 1.522809 1.263772246 1.781846
## 2020.423 1.548885 1.263872071 1.833899
## 2020.442 1.570498 1.267880783 1.873115
## 2020.462 1.636735 1.313109589 1.960360
## 2020.481 1.623589 1.283396830 1.963782
## 2020.500 1.641168 1.282917420 1.999420
## 2020.519 1.675402 1.301604052 2.049199
## 2020.538 1.672261 1.282348881 2.062174
## 2020.558 1.644330 1.239812446 2.048848
## 2020.577 1.579082 1.159834919 1.998330
## 2020.596 1.589457 1.156444783 2.022469
## 2020.615 1.596490 1.149796675 2.043183
## 2020.635 1.594098 1.134381421 2.053815
## 2020.654 1.557532 1.084965998 2.030097
## 2020.673 1.575469 1.090530508 2.060408
## 2020.692 1.601257 1.104153618 2.098361
## 2020.712 1.559131 1.050226882 2.068035
## 2020.731 1.543228 1.022736082 2.063719
## 2020.750 1.524066 0.992279902 2.055853
## 2020.769 1.480038 0.937162145 2.022914
## 2020.788 1.457599 0.903877762 2.011321
## 2020.808 1.484905 0.920530151 2.049280
## 2020.827 1.451998 0.877178961 2.026817
## 2020.846 1.449425 0.864339695 2.034511
## 2020.865 1.431333 0.836164730 2.026502
## 2020.885 1.424450 0.819361481 2.029538
## 2020.904 1.481809 0.866964993 2.096654
## 2020.923 1.531282 0.906830887 2.155732
## 2020.942 1.442919 0.809009138 2.076828
## 2020.962 1.376115 0.732884255 2.019346
## 2020.981 1.382902 0.730484521 2.035320
## 2021.000 1.375033 0.713555539 2.036511
## 2021.019 1.375736 0.705320513 2.046151
## 2021.038 1.313228 0.633992418 1.992463
## 2021.058 1.319250 0.631307952 2.007191
## 2021.077 1.359920 0.663379960 2.056460
## 2021.096 1.341913 0.636880002 2.046946
## 2021.115 1.360175 0.646749772 2.073600
## 2021.135 1.386827 0.665108002 2.108547
## 2021.154 1.406709 0.676789603 2.136629
## 2021.173 1.378643 0.640614868 2.116672
## 2021.192 1.424179 0.678129833 2.170229
## 2021.212 1.446175 0.692190034 2.200160
## 2021.231 1.451034 0.689195513 2.212872
## 2021.250 1.580730 0.811118718 2.350341
## 2021.269 1.647860 0.870554387 2.425166
## 2021.288 1.608278 0.823352169 2.393204
## 2021.308 1.565159 0.772687095 2.357631
## 2021.327 1.598186 0.798238454 2.398133
## 2021.346 1.593576 0.775318396 2.411833
## 2021.365 1.480450 0.642962839 2.317937
## 2021.385 1.478099 0.624153457 2.332044
## 2021.404 1.555287 0.687722835 2.422851
## 2021.423 1.606895 0.724058013 2.489731
## 2021.442 1.636773 0.740320949 2.533225
## 2021.462 1.707361 0.796461617 2.618261
## 2021.481 1.721435 0.797088350 2.645781
## 2021.500 1.738720 0.800543288 2.676896
## 2021.519 1.776382 0.825007236 2.727757
## 2021.538 1.762791 0.798077771 2.727505
## 2021.558 1.715558 0.737927199 2.693188
## 2021.577 1.605345 0.614786950 2.595903
## 2021.596 1.616621 0.613436120 2.619806
## 2021.615 1.614817 0.599062084 2.630573
## 2021.635 1.586503 0.558405656 2.614599
## 2021.654 1.513521 0.473172977 2.553869
## 2021.673 1.530468 0.478053033 2.582882
## 2021.692 1.552313 0.487937022 2.616689
## 2021.712 1.479298 0.403117452 2.555479
## 2021.731 1.461173 0.373297919 2.549048
## 2021.750 1.434526 0.335093730 2.533957
## 2021.769 1.380830 0.269951747 2.491708
## 2021.788 1.363531 0.241330265 2.485731
## 2021.808 1.448245 0.314830154 2.581660
## 2021.827 1.416111 0.271595532 2.560627
## 2021.846 1.426987 0.271474447 2.582500
## 2021.865 1.394300 0.227896366 2.560705
## 2021.885 1.378696 0.201499316 2.555892
## 2021.904 1.443413 0.255524470 2.631302
## 2021.923 1.480516 0.282028702 2.679003
## 2021.942 1.365351 0.156358523 2.574343
## 2021.962 1.317505 0.098097904 2.536912
## 2021.981 1.321515 0.091781581 2.551248
## 2022.000 1.295013 0.055039093 2.534986
## 2022.019 1.329384 0.079254568 2.579514
## 2022.038 1.279788 0.019583265 2.539992
## 2022.058 1.290278 0.020078556 2.560477
## 2022.077 1.319234 0.039118283 2.599350
## 2022.096 1.297032 0.007076133 2.586989
## 2022.115 1.347274 0.047552007 2.646997
## 2022.135 1.388800 0.079384878 2.698216
## 2022.154 1.434560 0.115523040 2.753598
## 2022.173 1.394999 0.066409815 2.723589Dataframe del pronostico
pronostico <- as.data.frame(forecast_mod)
head(pronostico,20)## Point Forecast Lo 95 Hi 95
## 2020.346 1.577947 1.440215 1.715679
## 2020.365 1.496456 1.291696 1.701216
## 2020.385 1.488244 1.247726 1.728763
## 2020.404 1.522809 1.263772 1.781846
## 2020.423 1.548885 1.263872 1.833899
## 2020.442 1.570498 1.267881 1.873115
## 2020.462 1.636735 1.313110 1.960360
## 2020.481 1.623589 1.283397 1.963782
## 2020.500 1.641168 1.282917 1.999420
## 2020.519 1.675402 1.301604 2.049199
## 2020.538 1.672261 1.282349 2.062174
## 2020.558 1.644330 1.239812 2.048848
## 2020.577 1.579082 1.159835 1.998330
## 2020.596 1.589457 1.156445 2.022469
## 2020.615 1.596490 1.149797 2.043183
## 2020.635 1.594098 1.134381 2.053815
## 2020.654 1.557532 1.084966 2.030097
## 2020.673 1.575469 1.090531 2.060408
## 2020.692 1.601257 1.104154 2.098361
## 2020.712 1.559131 1.050227 2.0680352.2 PREDICCION VOLUMEN DE VENTAS PARA AGUACATE TIPO ORGANICO
arima_vol_org <- auto.arima(organic_avgvol,d=1,D=1, trace=T, stepwise=F, approximation=F)forecast_mod <- forecast(arima_vol_org, h=96,level=95)Plot de la prediccion
plot(forecast_mod)
plot con el ajuste de la prediccion.
autoplot(forecast_mod)+autolayer(fitted(forecast_mod), series="Ajuste")
Sumario resultados
print(summary(forecast_mod))##
## Forecast method: ARIMA(0,1,1)(1,1,1)[52]
##
## Model Information:
## Series: organic_avgvol
## ARIMA(0,1,1)(1,1,1)[52]
##
## Coefficients:
## ma1 sar1 sma1
## -0.5292 -0.3664 -0.3371
## s.e. 0.0636 0.1434 0.1770
##
## sigma^2 = 58544490: log likelihood = -2342.95
## AIC=4693.9 AICc=4694.08 BIC=4707.56
##
## Error measures:
## ME RMSE MAE MPE MAPE MASE
## Training set 215.0373 6837.496 4508.321 -0.2774855 6.764423 0.2921328
## ACF1
## Training set 0.03536086
##
## Forecasts:
## Point Forecast Lo 95 Hi 95
## 2020.346 115025.07 100028.09 130022.0
## 2020.365 112699.48 96123.72 129275.2
## 2020.385 112337.76 94321.03 130354.5
## 2020.404 112449.65 93098.96 131800.3
## 2020.423 121154.55 100556.11 141753.0
## 2020.442 124543.78 102768.97 146318.6
## 2020.462 114196.84 91306.03 137087.6
## 2020.481 116308.21 92353.34 140263.1
## 2020.500 107389.09 82415.46 132362.7
## 2020.519 105249.82 79297.39 131202.3
## 2020.538 107287.20 80391.56 134182.8
## 2020.558 107132.65 79325.78 134939.5
## 2020.577 107821.80 79132.62 136511.0
## 2020.596 105180.53 75635.39 134725.7
## 2020.615 106167.87 75790.87 136544.9
## 2020.635 105200.47 74013.79 136387.1
## 2020.654 106320.44 74344.58 138296.3
## 2020.673 108536.49 75790.47 141282.5
## 2020.692 110827.99 77329.51 144326.5
## 2020.712 108929.88 74695.47 143164.3
## 2020.731 107019.88 72065.03 141974.7
## 2020.750 105477.32 69816.58 141138.1
## 2020.769 112311.43 75958.51 148664.4
## 2020.788 111745.92 74713.75 148778.1
## 2020.808 107624.32 69925.14 145323.5
## 2020.827 106603.78 68249.19 144958.4
## 2020.846 103478.55 64479.55 142477.5
## 2020.865 102137.59 62504.67 141770.5
## 2020.885 104336.42 64079.56 144593.3
## 2020.904 105333.66 64462.38 146204.9
## 2020.923 98466.75 56990.15 139943.3
## 2020.942 102899.53 60826.32 144972.7
## 2020.962 115621.50 72960.02 158283.0
## 2020.981 118343.07 75101.33 161584.8
## 2021.000 121081.35 77267.06 164895.6
## 2021.019 117307.54 72928.09 161687.0
## 2021.038 120073.82 75136.31 165011.3
## 2021.058 119385.87 73897.15 164874.6
## 2021.077 111065.88 65032.55 157099.2
## 2021.096 118171.73 71600.15 164743.3
## 2021.115 125502.45 78398.78 172606.1
## 2021.135 119657.45 72027.63 167287.3
## 2021.154 131635.80 83485.58 179786.0
## 2021.173 122227.96 73562.90 170893.0
## 2021.192 121214.36 72039.86 170388.9
## 2021.212 126582.00 76903.27 176260.7
## 2021.231 134261.87 84083.99 184439.8
## 2021.250 122761.09 72088.96 173433.2
## 2021.269 118648.99 67487.40 169810.6
## 2021.288 119488.23 67841.81 171134.6
## 2021.308 125282.19 73155.46 177408.9
## 2021.327 121216.65 68613.99 173819.3
## 2021.346 125762.87 71916.14 179609.6
## 2021.365 126129.06 71509.79 180748.3
## 2021.385 124326.33 68945.30 179707.4
## 2021.404 122745.78 66613.31 178878.2
## 2021.423 126588.76 69714.79 183462.7
## 2021.442 132213.38 74607.44 189819.3
## 2021.462 124053.70 65724.99 182382.4
## 2021.481 122473.74 63431.11 181516.4
## 2021.500 115941.33 56193.30 175689.4
## 2021.519 113863.88 53418.68 174309.1
## 2021.538 117112.54 55978.12 178247.0
## 2021.558 117207.86 55391.91 179023.8
## 2021.577 121582.89 59092.84 184072.9
## 2021.596 118205.10 55048.14 181362.1
## 2021.615 117640.90 53824.01 181457.8
## 2021.635 119456.20 54986.13 183926.3
## 2021.654 121591.52 56474.81 186708.2
## 2021.673 124122.70 58365.72 189879.7
## 2021.692 125067.17 58676.10 191458.2
## 2021.712 123540.55 56521.38 190559.7
## 2021.731 121643.90 54002.46 189285.3
## 2021.750 119514.82 51256.79 187772.9
## 2021.769 128837.06 59967.96 197706.2
## 2021.788 126801.06 57326.26 196275.9
## 2021.808 120064.51 49989.25 190139.8
## 2021.827 119892.37 49221.74 190563.0
## 2021.846 113914.24 42653.22 185175.2
## 2021.865 113455.85 41609.30 185302.4
## 2021.885 115547.57 43120.22 187974.9
## 2021.904 114428.80 41425.26 187432.3
## 2021.923 113571.89 39996.69 187147.1
## 2021.942 118545.90 44403.43 192688.4
## 2021.962 129928.04 55222.62 204633.5
## 2021.981 137077.77 61813.60 212341.9
## 2022.000 140667.54 64848.73 216486.4
## 2022.019 131667.40 55298.03 208036.8
## 2022.038 136982.68 60066.67 213898.7
## 2022.058 133274.87 55816.09 210733.7
## 2022.077 128157.47 50159.69 206155.2
## 2022.096 135131.41 56598.34 213664.5
## 2022.115 142197.13 63132.38 221261.9
## 2022.135 134395.30 54802.43 213988.2
## 2022.154 151216.18 71098.67 231333.7
## 2022.173 137871.98 57233.24 218510.7Dataframe del pronostico
pronostico <- as.data.frame(forecast_mod)
head(pronostico,20)## Point Forecast Lo 95 Hi 95
## 2020.346 115025.1 100028.09 130022.0
## 2020.365 112699.5 96123.72 129275.2
## 2020.385 112337.8 94321.03 130354.5
## 2020.404 112449.6 93098.96 131800.3
## 2020.423 121154.5 100556.11 141753.0
## 2020.442 124543.8 102768.97 146318.6
## 2020.462 114196.8 91306.03 137087.6
## 2020.481 116308.2 92353.34 140263.1
## 2020.500 107389.1 82415.46 132362.7
## 2020.519 105249.8 79297.39 131202.3
## 2020.538 107287.2 80391.56 134182.8
## 2020.558 107132.7 79325.78 134939.5
## 2020.577 107821.8 79132.62 136511.0
## 2020.596 105180.5 75635.39 134725.7
## 2020.615 106167.9 75790.87 136544.9
## 2020.635 105200.5 74013.79 136387.1
## 2020.654 106320.4 74344.58 138296.3
## 2020.673 108536.5 75790.47 141282.5
## 2020.692 110828.0 77329.51 144326.5
## 2020.712 108929.9 74695.47 143164.32.3 PREDICCION PRECIO PROMEDIO PARA AGUACATE TIPO CONVENCIONAL
arima_avg_con <- auto.arima(conventional_avgpr,d=1,D=1, trace=T, stepwise=F, approximation=F)forecast_mod <- forecast(arima_avg_con, h=96,level=95)Plot de la prediccion
plot(forecast_mod)
plot con el ajuste de la prediccion.
autoplot(forecast_mod)+autolayer(fitted(forecast_mod), series="Ajuste")
Sumario resultados
print(summary(forecast_mod))##
## Forecast method: ARIMA(2,1,0)(1,1,0)[52]
##
## Model Information:
## Series: conventional_avgpr
## ARIMA(2,1,0)(1,1,0)[52]
##
## Coefficients:
## ar1 ar2 sar1
## -0.0637 -0.1299 -0.3890
## s.e. 0.0660 0.0657 0.0621
##
## sigma^2 = 0.007142: log likelihood = 233.92
## AIC=-459.84 AICc=-459.66 BIC=-446.18
##
## Error measures:
## ME RMSE MAE MPE MAPE MASE
## Training set 0.0001134446 0.07551935 0.05096262 -0.1561508 4.490172 0.2863638
## ACF1
## Training set 0.002419695
##
## Forecasts:
## Point Forecast Lo 95 Hi 95
## 2020.346 1.1980365 1.03240146 1.363672
## 2020.365 1.0674377 0.84053532 1.294340
## 2020.385 1.1176644 0.85403148 1.381297
## 2020.404 1.1946086 0.89753600 1.491681
## 2020.423 1.2277776 0.89959596 1.555959
## 2020.442 1.2506181 0.89423126 1.607005
## 2020.462 1.2677212 0.88531103 1.650131
## 2020.481 1.2954900 0.88868887 1.702291
## 2020.500 1.3032944 0.87347351 1.733115
## 2020.519 1.2840478 0.83238294 1.735713
## 2020.538 1.2563611 0.78386205 1.728860
## 2020.558 1.2187499 0.72629674 1.711203
## 2020.577 1.1996607 0.68803098 1.711290
## 2020.596 1.1823876 0.65227462 1.712501
## 2020.615 1.1842049 0.63623182 1.732178
## 2020.635 1.1652413 0.59997214 1.730511
## 2020.654 1.1707584 0.58870685 1.752810
## 2020.673 1.2069861 0.60862267 1.805350
## 2020.692 1.1959111 0.58166884 1.810153
## 2020.712 1.1619179 0.53219709 1.791639
## 2020.731 1.1500515 0.50522356 1.794879
## 2020.750 1.1194155 0.45982635 1.779005
## 2020.769 1.0634763 0.38944918 1.737504
## 2020.788 1.0641115 0.37594915 1.752274
## 2020.808 1.0412312 0.33921823 1.743244
## 2020.827 1.0524395 0.33684396 1.768035
## 2020.846 1.0258273 0.29690228 1.754752
## 2020.865 1.0271673 0.28515220 1.769182
## 2020.885 0.9830441 0.22816581 1.737922
## 2020.904 1.0531850 0.28565914 1.820711
## 2020.923 1.1052834 0.32531499 1.885252
## 2020.942 1.0295779 0.23736234 1.821794
## 2020.962 0.9607918 0.15651556 1.765068
## 2020.981 0.9525726 0.13641386 1.768731
## 2021.000 0.9266939 0.09882324 1.754565
## 2021.019 1.0010743 0.16165508 1.840493
## 2021.038 0.8243621 -0.02644888 1.675173
## 2021.058 0.9261230 0.06407072 1.788175
## 2021.077 0.9954530 0.12230415 1.868602
## 2021.096 0.9953129 0.11120671 1.879419
## 2021.115 1.0474140 0.15248475 1.942343
## 2021.135 1.0491784 0.14355532 1.954802
## 2021.154 1.0682203 0.15202822 1.984412
## 2021.173 1.0855720 0.15893143 2.012213
## 2021.192 1.0852966 0.14832411 2.022269
## 2021.212 1.0534307 0.10623888 2.000622
## 2021.231 1.1352091 0.17790712 2.092511
## 2021.250 1.1974735 0.23016701 2.164780
## 2021.269 1.2260772 0.24886863 2.203286
## 2021.288 1.0841772 0.09716590 2.071189
## 2021.308 1.1304718 0.13375410 2.127189
## 2021.327 1.1844791 0.17814874 2.190810
## 2021.346 1.1991873 0.16464281 2.233732
## 2021.365 1.0833016 0.02273096 2.143872
## 2021.385 1.1130526 0.02974701 2.196358
## 2021.404 1.2017755 0.09587188 2.307679
## 2021.423 1.2388267 0.11047083 2.367182
## 2021.442 1.2618589 0.11154993 2.412168
## 2021.462 1.2937766 0.12196041 2.465593
## 2021.481 1.3152094 0.12226373 2.508155
## 2021.500 1.3351783 0.12146734 2.548889
## 2021.519 1.3125409 0.07841538 2.546666
## 2021.538 1.2753814 0.02117399 2.529589
## 2021.558 1.2237295 -0.05024354 2.497702
## 2021.577 1.1983066 -0.09513009 2.491743
## 2021.596 1.1685902 -0.14402150 2.481202
## 2021.615 1.1627846 -0.16872598 2.494295
## 2021.635 1.1468758 -0.20326917 2.497021
## 2021.654 1.1493101 -0.21921556 2.517836
## 2021.673 1.1697875 -0.21687520 2.556450
## 2021.692 1.1408324 -0.26373325 2.545398
## 2021.712 1.1084646 -0.31377850 2.530708
## 2021.731 1.1015027 -0.33820097 2.541206
## 2021.750 1.0687367 -0.38821818 2.525692
## 2021.769 1.0161887 -0.45781565 2.490193
## 2021.788 1.0199626 -0.47089609 2.510821
## 2021.808 0.9966900 -0.51083476 2.504215
## 2021.827 1.0026015 -0.52140699 2.526610
## 2021.846 0.9961395 -0.54417629 2.536455
## 2021.865 0.9945089 -0.56194345 2.550961
## 2021.885 0.9359246 -0.63649869 2.508348
## 2021.904 0.9977980 -0.59043557 2.586032
## 2021.923 1.0384901 -0.56539799 2.642378
## 2021.942 0.9459133 -0.67347791 2.565305
## 2021.962 0.9127474 -0.72199995 2.547495
## 2021.981 0.9152899 -0.73467068 2.565251
## 2022.000 0.8824049 -0.78262994 2.547440
## 2022.019 0.9594757 -0.72049811 2.639450
## 2022.038 0.7877521 -0.90702906 2.482533
## 2022.058 0.8994900 -0.80997021 2.608950
## 2022.077 0.9703053 -0.75370895 2.694320
## 2022.096 0.9656811 -0.77276538 2.704128
## 2022.115 1.0268345 -0.72592540 2.779594
## 2022.135 1.0345403 -0.73241710 2.801498
## 2022.154 1.0669222 -0.71411942 2.847964
## 2022.173 1.0773797 -0.71763576 2.872395Dataframe del pronostico
pronostico <- as.data.frame(forecast_mod)
head(pronostico,20)## Point Forecast Lo 95 Hi 95
## 2020.346 1.198037 1.0324015 1.363672
## 2020.365 1.067438 0.8405353 1.294340
## 2020.385 1.117664 0.8540315 1.381297
## 2020.404 1.194609 0.8975360 1.491681
## 2020.423 1.227778 0.8995960 1.555959
## 2020.442 1.250618 0.8942313 1.607005
## 2020.462 1.267721 0.8853110 1.650131
## 2020.481 1.295490 0.8886889 1.702291
## 2020.500 1.303294 0.8734735 1.733115
## 2020.519 1.284048 0.8323829 1.735713
## 2020.538 1.256361 0.7838621 1.728860
## 2020.558 1.218750 0.7262967 1.711203
## 2020.577 1.199661 0.6880310 1.711290
## 2020.596 1.182388 0.6522746 1.712501
## 2020.615 1.184205 0.6362318 1.732178
## 2020.635 1.165241 0.5999721 1.730511
## 2020.654 1.170758 0.5887068 1.752810
## 2020.673 1.206986 0.6086227 1.805350
## 2020.692 1.195911 0.5816688 1.810153
## 2020.712 1.161918 0.5321971 1.7916392.4 PREDICCION VOLUMEN VENTAS PARA AGUACATE TIPO CONVENCIONAL
arima_vol_con <- auto.arima(conventional_avgvol,d=1,D=1, trace=T, stepwise=F, approximation=F)forecast_mod <- forecast(arima_vol_con, h=96,level=95)Plot de la prediccion
plot(forecast_mod)
plot con el ajuste de la prediccion.
autoplot(forecast_mod)+autolayer(fitted(forecast_mod), series="Ajuste")
Sumario resultados
print(summary(forecast_mod))##
## Forecast method: ARIMA(0,1,2)(0,1,1)[52]
##
## Model Information:
## Series: conventional_avgvol
## ARIMA(0,1,2)(0,1,1)[52]
##
## Coefficients:
## ma1 ma2 sma1
## -0.5902 -0.2181 -0.5794
## s.e. 0.0680 0.0726 0.0948
##
## sigma^2 = 6.397e+10: log likelihood = -3127.96
## AIC=6263.92 AICc=6264.11 BIC=6277.59
##
## Error measures:
## ME RMSE MAE MPE MAPE MASE ACF1
## Training set 69.39602 226020 136484.4 -0.8917261 7.11636 0.4895279 -0.01003706
##
## Forecasts:
## Point Forecast Lo 95 Hi 95
## 2020.346 2617513 2121082 3113943
## 2020.365 2797144 2260644 3333644
## 2020.385 2567407 2022532 3112283
## 2020.404 2492944 1939819 3046069
## 2020.423 2521168 1959914 3082421
## 2020.442 2457490 1888225 3026756
## 2020.462 2506964 1929797 3084130
## 2020.481 2378913 1793953 2963873
## 2020.500 2404246 1811594 2996898
## 2020.519 2397956 1797711 2998201
## 2020.538 2382148 1774404 2989891
## 2020.558 2352795 1737645 2967945
## 2020.577 2327991 1705522 2950460
## 2020.596 2382526 1752824 3012229
## 2020.615 2385551 1748697 3022405
## 2020.635 2342212 1698286 2986138
## 2020.654 2330846 1679924 2981767
## 2020.673 2265688 1607846 2923531
## 2020.692 2273406 1608715 2938097
## 2020.712 2271148 1599678 2942618
## 2020.731 2215754 1537573 2893935
## 2020.750 2202944 1518117 2887770
## 2020.769 2245096 1553688 2936504
## 2020.788 2212872 1514944 2910799
## 2020.808 2222133 1517746 2926519
## 2020.827 2187875 1477087 2898662
## 2020.846 2200046 1482915 2917176
## 2020.865 2123964 1400545 2847382
## 2020.885 2272571 1542919 3002223
## 2020.904 2018768 1282935 2754601
## 2020.923 2034015 1292053 2775977
## 2020.942 2207413 1459372 2955453
## 2020.962 2516294 1762223 3270365
## 2020.981 2556209 1796166 3316253
## 2021.000 2735472 1969674 3501270
## 2021.019 2581068 1809390 3352747
## 2021.038 3200354 2422840 3977868
## 2021.058 2706158 1922852 3489464
## 2021.077 2613395 1824339 3402451
## 2021.096 2770138 1975374 3564903
## 2021.115 2765127 1964695 3565558
## 2021.135 2628823 1822764 3434882
## 2021.154 2797897 1986249 3609544
## 2021.173 2720630 1903432 3537829
## 2021.192 2647883 1825172 3470594
## 2021.212 2779358 1951171 3607545
## 2021.231 2694513 1860886 3528141
## 2021.250 2672920 1833888 3511953
## 2021.269 2662879 1818477 3507282
## 2021.288 3060352 2210612 3910091
## 2021.308 2943675 2088633 3798718
## 2021.327 2766813 1906500 3627126
## 2021.346 2869787 1957088 3782486
## 2021.365 3042671 2112192 3973149
## 2021.385 2812934 1872672 3753197
## 2021.404 2738471 1788525 3688418
## 2021.423 2766695 1807162 3726227
## 2021.442 2703017 1733994 3672040
## 2021.462 2752491 1774068 3730913
## 2021.481 2624440 1636708 3612171
## 2021.500 2649773 1652819 3646727
## 2021.519 2643483 1637391 3649575
## 2021.538 2627675 1612527 3642823
## 2021.558 2598322 1574198 3622446
## 2021.577 2573518 1540497 3606540
## 2021.596 2628053 1586210 3669896
## 2021.615 2631078 1580487 3681669
## 2021.635 2587739 1528473 3647006
## 2021.654 2576373 1508501 3644244
## 2021.673 2511215 1434808 3587623
## 2021.692 2518933 1434057 3603810
## 2021.712 2516675 1423395 3609955
## 2021.731 2461281 1359662 3562900
## 2021.750 2448471 1338575 3558367
## 2021.769 2490623 1372512 3608734
## 2021.788 2458399 1332132 3584665
## 2021.808 2467660 1333296 3602023
## 2021.827 2433402 1290999 3575805
## 2021.846 2445573 1295186 3595959
## 2021.865 2369491 1211176 3527805
## 2021.885 2518098 1351909 3684286
## 2021.904 2264295 1090285 3438305
## 2021.923 2279542 1097762 3461322
## 2021.942 2452939 1263440 3642439
## 2021.962 2761821 1564652 3958989
## 2021.981 2801736 1596937 4006536
## 2022.000 2980999 1768677 4193321
## 2022.019 2826595 1606768 4046422
## 2022.038 3445881 2218596 4673167
## 2022.058 2951685 1716985 4186384
## 2022.077 2858922 1616853 4100991
## 2022.096 3015665 1766270 4265061
## 2022.115 3010654 1753975 4267332
## 2022.135 2874350 1610430 4138270
## 2022.154 3043424 1772304 4314544
## 2022.173 2966157 1687878 4244437Dataframe del pronostico
pronostico <- as.data.frame(forecast_mod)
head(pronostico,20)## Point Forecast Lo 95 Hi 95
## 2020.346 2617513 2121082 3113943
## 2020.365 2797144 2260644 3333644
## 2020.385 2567407 2022532 3112283
## 2020.404 2492944 1939819 3046069
## 2020.423 2521168 1959914 3082421
## 2020.442 2457490 1888225 3026756
## 2020.462 2506964 1929797 3084130
## 2020.481 2378913 1793953 2963873
## 2020.500 2404246 1811594 2996898
## 2020.519 2397956 1797711 2998201
## 2020.538 2382148 1774404 2989891
## 2020.558 2352795 1737645 2967945
## 2020.577 2327991 1705522 2950460
## 2020.596 2382526 1752824 3012229
## 2020.615 2385551 1748697 3022405
## 2020.635 2342212 1698286 2986138
## 2020.654 2330846 1679924 2981767
## 2020.673 2265688 1607846 2923531
## 2020.692 2273406 1608715 2938097
## 2020.712 2271148 1599678 2942618