Ejercicio Arima

Ejercicio ARIMA con Google (GOOG)

install.packages('quantmod') # librería para hacer scraping

## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.3'
## (as 'lib' is unspecified)

install.packages('tseries')

## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.3'
## (as 'lib' is unspecified)

install.packages('timeSeries')

## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.3'
## (as 'lib' is unspecified)

install.packages('forecast')

## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.3'
## (as 'lib' is unspecified)

install.packages('xts')

## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.3'
## (as 'lib' is unspecified)

install.packages('ggplot2')

## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.3'
## (as 'lib' is unspecified)

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)

1. Identificación de datos

GOOG <- getSymbols('GOOG', src='yahoo', from = as.Date("2019-01-01"),to=as.Date("2023-08-04"), auto.assign = FALSE)

Graficando la serie

chartSeries(GOOG, name="GOOG", subset="last 6 months", theme=chartTheme("white"))

Datos de yahoo finance:

data_GOOG <- data.frame(GOOG, tiempo = as.Date(rownames(data.frame(GOOG))))
head(data_GOOG)

##            GOOG.Open GOOG.High GOOG.Low GOOG.Close GOOG.Volume GOOG.Adjusted
## 2019-01-02   50.8285   52.6160  50.7855    52.2925    30652000       52.2925
## 2019-01-03   52.0500   52.8490  50.7035    50.8030    36822000       50.8030
## 2019-01-04   51.6295   53.5420  51.3709    53.5355    41878000       53.5355
## 2019-01-07   53.5750   53.7000  52.7380    53.4195    39638000       53.4195
## 2019-01-08   53.8055   54.2280  53.0265    53.8140    35298000       53.8140
## 2019-01-09   54.0825   54.1315  53.3200    53.7330    23986000       53.7330
##                tiempo
## 2019-01-02 2019-01-02
## 2019-01-03 2019-01-03
## 2019-01-04 2019-01-04
## 2019-01-07 2019-01-07
## 2019-01-08 2019-01-08
## 2019-01-09 2019-01-09

attach(data_GOOG)

Separando la serie close:

base1 = data.frame(tiempo, GOOG.Close)
names (base1) = c("tiempo","GOOG")
base1 <- na.omit(base1) #eliminando datos ominitidos "NA"
#base1
head(base1, n = 10)

##        tiempo    GOOG
## 1  2019-01-02 52.2925
## 2  2019-01-03 50.8030
## 3  2019-01-04 53.5355
## 4  2019-01-07 53.4195
## 5  2019-01-08 53.8140
## 6  2019-01-09 53.7330
## 7  2019-01-10 53.5165
## 8  2019-01-11 52.8595
## 9  2019-01-14 52.2345
## 10 2019-01-15 53.8575

Graficando la serie

ggplot(base1, aes(x = tiempo, y = GOOG)) + geom_line() +  labs(title = "Cotizaciones de Google", x = "Fecha", y = "Precios por accion")

SEGUNDA ETAPA: Análisis y diferenciación de la serie temporal

GOOG_ma = ts(na.omit(base1$GOOG), frequency=30)
decomp = stl(GOOG_ma, s.window="periodic")
deseasonal_base1 <- seasadj(decomp)
plot(decomp)

adf.test(GOOG_ma, alternative = "stationary")

## 
##  Augmented Dickey-Fuller Test
## 
## data:  GOOG_ma
## Dickey-Fuller = -1.4928, Lag order = 10, p-value = 0.793
## alternative hypothesis: stationary

p-value = 0.793 > 0.05 NRHo no existe estacionariedad

TERCERA ETAPA: Ajuste de un modelo ARIMA.

Ajustamos el modelo.

auto.arima(deseasonal_base1, seasonal=FALSE)

## Series: deseasonal_base1 
## ARIMA(0,1,0) 
## 
## sigma^2 = 3.737:  log likelihood = -2398.07
## AIC=4798.14   AICc=4798.14   BIC=4803.19