If you find exercises done at the lecture and seminar as well as those presented in this file not

sufficient, study Chapter 14 “Time Series“ in Statistics for Business and Economics by McClave,
Benson and Sincich. (Practise only the examples on the topics, methods and tools introduced in the
Lecture6 presentation. If in doubt whether a particular exercise belongs to 4ST601 course, contact
your seminar teacher.)

Exercise ts_mov1) “Exercise_ts_mov1“ sheet contains data on the stock of cattle (in thousands of
heads) in the Czech Republic in the period 1993–2006. Smooth the time series with simple moving
averages of the lengths 3, 5 and 7.

/e.g. for m = 3:

e.g. for m = 5:

e.g. for m = 7: always in thousands of heads/

Exercise ts_mov2) In the table below, there are data about loans provided by a bank (in millions of
CZK) in eleven months of the year 2005. Smooth the time series with moving averages of the length
4.
Month yt Month yt
January 50 July 32.6
February 36.5 August 38.7
March 43 September 41.7
April 44.5 October 41.1
May 38.9 November 33.8
June 38.1
/e.g. millions of CZK/

Exercise ts_mov3) The “Exercise_ts_mov3“ sheet contains the numbers of cinema-goers on chosen
days over five successive weeks. Smooth out the time series with appropriate moving averages. Draw
a time series plot with the original time series and the series of moving averages (i.e. the smoothed
time series).

/e.g. cinema-goers/

Exercise ts_mov4) In the table below, there are quarterly data about frozen dairy product values (in
millions of CZK, comparable prices) in the years 2002–2005. Smooth the time series with appropriate
moving averages.
Year Quarter yt Year Quarter yt
2002 1 35 2004 1 34
2 46 2 61
3 71 3 94
4 44 4 33
2003 1 39 2005 1 54
2 76 2 70
3 84 3 102
4 50 4 48

/e.g. millions of CZK/

Exercise ts_mov5) There are monthly data on total construction works (in millions of CZK, regural
prices) in the Czech Republic over the years 1994–2006 on the “Exercise_ts_mov5“ sheet. Smooth
the time series with appropriate moving averages. Draw a time series plot with the original time
series and the series of moving averages (i.e. the smoothed time series).

/e.g. millions of CZK /

Exercise ts_mov6) We have the following time series of the numbers of guests of a theme park (in
thousands) available:
Quarter 2009 2010 2011
1. 20 21 24
2. 12 11 12
3. 36 44 48
4. 30 35 40
Smooth the time series of the numbers of guests of the seasonal component by appropriate moving
averages. Calculate only the first five values.

/ guests/

Exercise ts_mov7) Smooth the following time series from the seasonal component using appropriate
moving averages. Calculate and write down only the first five moving averages, indicating the time
point to which each moving average corresponds.

Week
Day
1. 2. 3.
Monday 6 8 7
Tuesday 7 10 9
Wednesda 8 11 10
y
Thursday 9 9 12
Friday 8 7 9
Saturday 2 3 4
Sunday 1 1 2

/ /

Exercise ts_mov8) We know the numbers of visitors in a small cinema on particular days over a
three-week period.
Day
Week Monda Tuesday Wednesda Thursday Friday Saturday Sunday
y y
1. 54 55 42 48 78 76 49
2. 42 44 55 43 58 76 59
3. 45 38 46 41 60 66 53
Smooth out this time series of the seasonal component with appropriate moving averages.

/e.g. visitors/

Exercise ts_seas1) Use the results of the exercise “ts_mov4“ to calculate the strength of the seasonal
component by a simple approach:
a) assuming additive decomposition,
b) assuming multiplicative decomposition.

/a) e.g. , b) e.g. /

Exercise ts_seas2) Use the results of the exercise “ts_mov3“ to calculate the strength of the seasonal
component by a simple approach:
a) assuming additive decomposition,
b) assuming multiplicative decomposition.

/a) e.g. Monday: visitors, Saturday: visitors,

b) e.g. Wednesday: /

Exercise ts_seas3) Use the results of the exercise “ts_mov5“ to calculate the strength of the seasonal
component by a simple approach:
a) assuming additive decomposition,
b) assuming multiplicative decomposition.

/a) e.g. January: millions of CZK, June: millions of CZK

b) e.g. December: /

Exercise ts_seas4) There is a quarterly time series of the Czech Republic GDP expenditure (in millions
of CZK, regular prices) in the period 2007Q1–2012Q2 on the “Exercise_ts_seas4“ sheet. Draw a time
series plot for this time series. Calculate trend and seasonal components applying a regression
approach and use the constructed function to predict GDP in the period 2012Q3–2013Q2.
/

e.g. millions of CZK/

Exercise ts_seas5) A quarterly time series of visitors’ numbers (in thousands) in a cultural centre
from the 1st quarter of 2012 to the 4th quarter of 2014 was analysed, the following regression
function was constructed, using a regression approach to the seasonal component:

a) Write down the formula of the constructed linear trend function.

b) Calculate seasonal differences for individual quarters. In what units are these differences
expressed?
c) Predict visitors’ numbers for the 1st quarter of the year 2015.

/b) e.g. thousands of visitors,

c) thousands of visitors /