La base de datos para el siguiente análsisi es extraÃdo de Yahoo Finance (a través de la técnica scraping) y trata sobre el precio de las acciones de McDonalds.
#install.packages('quantmod') # librerÃa para hacer scraping
#install.packages('tseries')
#install.packages('timeSeries')
#install.packages('forecast')
#install.packages('xts')
#install.packages('ggplot2')
library(quantmod)
## Loading required package: xts
## Loading required package: zoo
##
## Attaching package: 'zoo'
## The following objects are masked from 'package:base':
##
## as.Date, as.Date.numeric
## Loading required package: TTR
## Registered S3 method overwritten by 'quantmod':
## method from
## as.zoo.data.frame zoo
library(tseries)
library(timeSeries)
## Loading required package: timeDate
##
## Attaching package: 'timeSeries'
## The following object is masked from 'package:zoo':
##
## time<-
library(forecast)
library(xts)
library(ggplot2)
MCD <- getSymbols('MCD', src='yahoo', from = as.Date("2020-01-01"),to=as.Date("2022-01-01"), auto.assign = FALSE)
Graficando la serie
chartSeries(MCD, name="MCD", subset="last 6 months", theme=chartTheme("white"))
Datos de yahoo finance:
data_MCD <- data.frame(MCD, tiempo = as.Date(rownames(data.frame(MCD))))
head(data_MCD)
## MCD.Open MCD.High MCD.Low MCD.Close MCD.Volume MCD.Adjusted
## 2020-01-02 198.00 200.80 197.81 200.79 3554200 185.2633
## 2020-01-03 199.39 200.55 198.85 200.08 2767600 184.6082
## 2020-01-06 199.60 202.77 199.35 202.33 4660400 186.6842
## 2020-01-07 201.87 202.68 200.51 202.63 4047400 186.9610
## 2020-01-08 202.62 206.69 202.20 205.91 5284200 189.9874
## 2020-01-09 206.86 209.37 206.10 208.35 5971600 192.2387
## tiempo
## 2020-01-02 2020-01-02
## 2020-01-03 2020-01-03
## 2020-01-06 2020-01-06
## 2020-01-07 2020-01-07
## 2020-01-08 2020-01-08
## 2020-01-09 2020-01-09
attach(data_MCD)
Separando la serie close:
base1 = data.frame(tiempo, MCD.Close)
names (base1) = c("tiempo","MCD")
base1 <- na.omit(base1) #eliminando datos ominitidos "NA"
#base1
head(base1, n = 10)
## tiempo MCD
## 1 2020-01-02 200.79
## 2 2020-01-03 200.08
## 3 2020-01-06 202.33
## 4 2020-01-07 202.63
## 5 2020-01-08 205.91
## 6 2020-01-09 208.35
## 7 2020-01-10 207.27
## 8 2020-01-13 206.51
## 9 2020-01-14 207.32
## 10 2020-01-15 209.77
Graficando la serie
ggplot(base1, aes(x = tiempo, y = MCD)) + geom_line() + geom_smooth(se = FALSE)+ labs(title = "Precio de las acciones de McDonalds", x = "Fecha", y = "Precio / Acción")
## `geom_smooth()` using method = 'loess' and formula = 'y ~ x'
MCD_ma = ts(na.omit(base1$MCD), frequency=30)
decomp = stl(MCD_ma, s.window="periodic")
deseasonal_base1 <- seasadj(decomp)
plot(decomp)
adf.test(MCD_ma, alternative = "stationary")
##
## Augmented Dickey-Fuller Test
##
## data: MCD_ma
## Dickey-Fuller = -3.1901, Lag order = 7, p-value = 0.08964
## alternative hypothesis: stationary
Contraste de hipótesis:
pp.test(MCD_ma, alternative = "stationary")
##
## Phillips-Perron Unit Root Test
##
## data: MCD_ma
## Dickey-Fuller Z(alpha) = -19.381, Truncation lag parameter = 5, p-value
## = 0.08117
## alternative hypothesis: stationary
Contraste de hipótesis:
Las ACF proporcionan información sobre cómo una observación influye en las siguientes.
Acf(MCD_ma, main='')
Pacf(MCD_ma, main='')
Para realizar un modelo ARIMA, la serie temporal debe ser estacionaria. Para conseguir esta estacionariedad, la diferenciaremos.
MCD_d1 = diff(deseasonal_base1, differences = 1)
plot(MCD_d1)
Para comprobar que la serie es, efectivamente, estacionaria, hacemos de nuevo el test aumentado de Dickey-Fuller.
adf.test(MCD_d1, alternative = "stationary")
## Warning in adf.test(MCD_d1, alternative = "stationary"): p-value smaller than
## printed p-value
##
## Augmented Dickey-Fuller Test
##
## data: MCD_d1
## Dickey-Fuller = -8.4103, Lag order = 7, p-value = 0.01
## alternative hypothesis: stationary
modeloarima<-auto.arima(MCD_ma, seasonal=FALSE)
modeloarima
## Series: MCD_ma
## ARIMA(2,1,3) with drift
##
## Coefficients:
## ar1 ar2 ma1 ma2 ma3 drift
## -1.7013 -0.8612 1.6012 0.7286 0.0178 0.1339
## s.e. 0.0553 0.0477 0.0720 0.0842 0.0483 0.1366
##
## sigma^2 = 10.78: log likelihood = -1311.7
## AIC=2637.41 AICc=2637.63 BIC=2666.96
tsdisplay(residuals(modeloarima), lag.max=10, main='(2,1,3) Model Residuals')
prediccion <- forecast(modeloarima, h=30)
plot(prediccion)
tail(prediccion$mean,30)
## Time Series:
## Start = c(17, 26)
## End = c(18, 25)
## Frequency = 30
## [1] 268.2279 268.1453 268.6385 268.3476 268.8947 268.6914 269.0431 269.0968
## [9] 269.1796 269.4696 269.3820 269.7583 269.6705 269.9727 270.0111 270.1625
## [17] 270.3489 270.3785 270.6446 270.6433 270.8934 270.9461 271.1180 271.2571
## [25] 271.3494 271.5496 271.6066 271.8143 271.8888 272.0601