Welcome to the co u rse ! TIME SE R IE S AN ALYSIS IN R Da v id S - - PowerPoint PPT Presentation

▶

Nov 04, 2022 357 likes •570 views

Welcome to the co u rse ! TIME SE R IE S AN ALYSIS IN R Da v id S . Ma eson Associate Professor at Cornell Uni v ersit y Introd u ction Time Series : A seq u ence of data in chronological order . Data is commonl y recorded seq u entiall y, o

SLIDE 1

Welcome to the course!

TIME SE R IE S AN ALYSIS IN R

David S. Maeson

Associate Professor at Cornell University

SLIDE 2

TIME SERIES ANALYSIS IN R

Introduction

Time Series: A sequence of data in chronological order. Data is commonly recorded sequentially, over time. Time series data is everywhere.

SLIDE 3

TIME SERIES ANALYSIS IN R

Time series example

Monthly values of the Consumer Price Index (CPI):

SLIDE 4

TIME SERIES ANALYSIS IN R

Time series data

Time series data is dated or time stamped in R.

print(BMW_data) ... 1996-07-08 0.002 1996-07-09 -0.006 1996-07-10 -0.016 1996-07-11 -0.020 1996-07-14 -0.006 1996-07-15 -0.014 1996-07-16 0.002 1996-07-17 -0.001 ...

SLIDE 5

TIME SERIES ANALYSIS IN R

Time series plots

plot(Time_Series)

SLIDE 6

TIME SERIES ANALYSIS IN R

Basic time series models

White Noise (WN) Random Walk (RW) Autoregression (AR) Simple Moving Average (MA)

Throughout this course, you will not only be learning how to use R for time series analysis and forecasting, you will also learn several models for time

SLIDE 7

Time series plots

TIME SE R IE S AN ALYSIS IN R

SLIDE 8

Sampling frequency

TIME SE R IE S AN ALYSIS IN R

David S. Maeson

Associate Professor at Cornell University

SLIDE 9

TIME SERIES ANALYSIS IN R

Sampling frequency: exact

Some time series data is exactly evenly spaced.

SLIDE 10

TIME SERIES ANALYSIS IN R

Sampling frequency: approximate

Some time series data is only approximately evenly spaced.

SLIDE 11

TIME SERIES ANALYSIS IN R

Sampling frequency: missing values

Some time series data is evenly spaced, but with missing values.

SLIDE 12

TIME SERIES ANALYSIS IN R

Basic assumptions

Simplifying assumptions for time series: Consecutive observations are equally spaced. Apply a discrete-time observation index. This may only hold approximately.

Ex. Daily log returns on stock may only be available for

weekdays.

Ex. Monthly CPI values are equally spaced by month, not by

days.

SLIDE 13

TIME SERIES ANALYSIS IN R

Sampling frequency: R functions

R functions: start() ,

end() , frequency() , deltat() start(Hourly_series) 1 1 end(Hourly_series) 1 24 frequency(Hourly_series) 24 deltat(Hourly_series) 0.0417

SLIDE 14

Let's practice!

TIME SE R IE S AN ALYSIS IN R

SLIDE 15

Basic time series

bjects

TIME SE R IE S AN ALYSIS IN R

David S. Maeson

Associate Professor at Cornell University

SLIDE 16

TIME SERIES ANALYSIS IN R

Building ts() objects - I

Start with a vector of data Apply the ts() function

data_vector 10 6 11 8 10 3 6 9 time_series <- ts(data_vector) plot(time_series)

SLIDE 17

TIME SERIES ANALYSIS IN R

Building ts() objects - II

Specify the start date and observation frequency:

time_series <- ts(data_vector, start = 2001, frequency =

plot(time_series)

SLIDE 18

TIME SERIES ANALYSIS IN R

Using is.ts()

The is.ts() function checks whether an object is of the

ts() class: is.ts(data_vector) FALSE is.ts(time_series) TRUE

SLIDE 19

TIME SERIES ANALYSIS IN R

Why ts() objects?

Why create and use time series objects of the ts() class? Improved ploing. Access to time index information. Model estimation and forecasting (later chapters).

SLIDE 20

Let's practice!

TIME SE R IE S AN ALYSIS IN R