Index

A

additive outliers

about 115

Danish fertility example 139–140

Danish hotel overnight stays by US citizens example 121–122

e-commerce example 117

aggregation of time series

about 19, 161

aggregation to monthly averages 168–173

aggregation to series observed every second day 179–181

aggregation to weekly averages 173–179

aggregation to yearly averages 161–168

analysis of daily series 181–185

TIMESERIES procedure supporting 19–22

AIC (Akaike information criterion) 77

Airline Model 80

Akaike information criterion (AIC) 77

ALPHA= option

FORECAST statement (ESM) 45

FORECAST statement (UCM) 138, 163

ANYDTDTE informat 12–13

ANYDTDTE20. informat 13, 16

ANYDTDTM informat 13

ANYDTTME informat 13

Arctic ice coverage example

about 185–186

aggregation to monthly averages 168–173

aggregation to series observed every second day 179–181

aggregation to weekly averages 173–179

aggregation to yearly averages 161–168

analysis of daily series 181–185

ARIMA (Box-Jenkins) models

about 79–82

seasonal adjustments and 116–118, 123–127

unobserved components models and 132, 134

ARIMA procedure

about 5

ARIMA (Box-Jenkins) models and 81

ARMA models and 172

ARIMA statement, X12 procedure 118

ARMA models

Arctic ice coverage example 172

Danish fertility example 82–85

Swiss Business Indicator example 90–95

autocorrelations 78, 81

AUTOMDL statement, X12 procedure 117–118, 123

AUTOREG procedure

about 5

ARIMA (Box-Jenkins) models and 81

ARMA models and 172

AUTOREG statement, UCM procedure 164

autoregressive models

about 76–77

Arctic ice coverage example 171–173

exponential smoothing and 75–76

fitting 77–78

B

BACK= option, ESM procedure 34, 66, 86–87

Box-Cox transformations 116

Box-Jenkins ARIMA models

about 79–82

seasonal adjustments and 116–118, 123–127

unobserved components models and 132, 134

C

Census Bureau X11 method

See X11 method

confidence limits

applying for forecasts 42–43

calculating for forecasts 42

Danish fertility example 43–46

continuous time series 4

CONVERT statement, EXPAND procedure

METHOD= option 23

TRANSFORMOUT= option 25

CYCLE statement, UCM procedure 145, 176–177

cycles, trend and business 101–103, 124–125

D

damped-trend method 49–51

Danish fertility example

component estimation 137–139

confidence limits for forecasts 43–46

estimating smoothing parameter with ESM procedure 47–49

forecast errors 38–40

forecasting by exponential smoothing 33–38

forecasting using damped-trend method 50–51

forecasting with fitted ARMA models 82–85

model extensions 142–148

outlier detection 139–142

UCM procedure and 137–148

Danish hotel overnight stays by US citizens example

exponential smoothing and 57–61

seasonal adjustment and 121–128

data sets, datetime variables and 16–17

date formats in procedure calls 10–11

DATEPART function 11, 14

DATESTYLE system option 13, 16

datetime variables

about 9

calculating time span between dates 17

extracting data from date variable example 15

handling 14–16

importing 12–14

output formats 9–12

datetime variables (continued)

plotting quarterly data by year example 14–15

time series data sets and 16–17

DATETIME20. format 11

DATETIME20.3 format 12

deleting outliers 185

DEPLAG statement, UCM procedure

about 177

ARMA models and 172

LAG= option 146–147

DIF function 16–17, 20

DIF= option, MODEL statement (VARMAX) 84

DIF12 function 16–17

discrete time series 4

dollar sign ($) 9

double exponential smoothing

about 30–33

Danish fertility example 36, 45

forecasting variance for 42

E

e-commerce example

ESM procedure supporting 63–74

estimation of components 149–155

model fit 158–160

regression components 155–158

seasonal adjustment in 103–108, 116–120

UCM procedure and 149–160

Winters method 63–68

X11 method and 103–108

X12 procedure and 116–120

ESM procedure

See also FORECAST statement, ESM procedure

about 5

BACK= option 34, 66, 86–87

damped-trend method 50–51

Danish fertility example 33–36, 43–53

Danish hotel overnight stays by US citizens example 59–63

e-commerce example 63–74

estimating smoothing parameter 47–49

ID statement 64

LEAD= option 65

OUTFOR= option 33–34

PLOT= option 44

PRINT= option 33–34, 47, 51, 66, 87

Swiss Business Indicator example 86–90

transformation facilities in 70–74

UCM procedure and 147, 160

Winters method 63–68

X12 procedure and 116, 128

ESTIMATE statement, UCM procedure

about 146, 158, 164

EXTRADIFFUSE= option 170

PLOT= option 139, 141, 146, 158, 161, 167

EURDFDD10. format 11

EURDFMY5. format 10

EURDFMY7. format 10

EXPAND procedure

about 5, 23

Arctic ice coverage example 181–182, 185

CONVERT statement 23, 25

time series aggregation example 22

TO= option 23

exponential smoothing

about 29–30, 53, 55–56

additive smoothing 61–63

applying confidence limits for forecasts 42–43

basic formula 40–41

calculating confidence limits for forecasts 42

damped-trend method 49–53

Danish fertility example 33–40, 44

Danish hotel overnight stays by US citizens example 57–61

determining smoothing constant 46–47

double 30–33, 36, 42

ESM procedure supporting 61–63, 67–74

estimating smoothing parameter in ESM procedure 47–49

Holt method 49–50

moving averages and 40–41

nonseasonal 29–53

parameterized models versus 75–95

relative importance of e-commerce example
67–74

seasonal 55–74

US retail e-commerce example 63–67

Winters method 55–56, 63–68

EXTRADIFFUSE= option, ESTIMATE statement (UCM) 170

F

fertility example

See Danish fertility example

FIRSTOBS= option, PRINT procedure 12

FORCE= option, X11 statement (X12) 123

forecast errors

See prediction errors

FORECAST statement, ESM procedure

ALPHA= option 45

MEDIAN option 72

METHOD= option 50–51, 59, 61, 64, 87

MODEL= option 35, 44

TRANSFORM= option 74

FORECAST statement, UCM procedure

about 138, 163–164

ALPHA= option 138, 163

forecasts

applying confidence limits for 42–43

calculating confidence limits for 42

confidence limits for Danish fertility example
43–46

exponential smoothing of nonseasonal series
29–53

exponential smoothing of seasonal series 55–74

exponential smoothing versus parameterized models 75–95

X12 procedure and 115–116

formats, including in procedure calls 10–12

G

Gaussian distribution

Danish fertility example 39

unobserved components models 135

H

Henderson moving average 124

Holt exponential smoothing method 49–50

I

ID statement

ESM procedure 64

TIMESERIES procedure 168, 179

UCM procedure 137, 161, 168, 181

VARMAX procedure 82

importing datetime variables 12–14

INTCK function 17

interpolation of time series 23–26

INTERVAL= option

ID statement (ESM) 64

ID statement (TIMESERIES) 168, 179

ID statement (UCM) 137, 168, 181

ID statement (VARMAX) 82

IRREGULAR statement, UCM procedure 172, 182

K

Kalman filter 134–135

KEEPH= option, SEASON statement (UCM) 175

L

LAG= option, DEPLAG statement (UCM) 146–147

LAG12 function 16–17

LEAD= option

ESM procedure 65

OUTPUT statement (VARMAX) 91

LEVEL statement, UCM procedure

about 149, 157

PLOT= option 138, 142

PRINT= option 142

likelihood function 81, 135

likelihood ratio testing 81

Ljung-Box test 118

M

maximum likelihood estimation 81

MEDIAN option, FORECAST statement (ESM) 72

METHOD= option

CONVERT statement (EXPAND) 23

FORECAST statement (ESM) 50–51, 59, 61, 64, 87

MODEL statement (VARMAX) 91

MINIC= option, MODEL statement (VARMAX) 91

missing values

autoregressive models and 76

EXPAND procedure and 23

TIMESERIES procedure and 21–22, 181–182

UCM procedure and 185

MMDDYY6. format 10

MMDDYY10. format 10

MODE= option, X11 statement (X12) 121

model fitting

ARMA models 82–85

autoregressive models 77–78

X12 procedure and 115–116

MODEL= option, FORECAST statement (ESM) 35, 44

MODEL statement, UCM procedure 137, 156

MODEL statement, VARMAX procedure

about 82

DIF= option 84

METHOD= option 91

MINIC= option 91

NOINT option 84

NSEASON= option 94

PRINT= option 94

TREND= option 94

monthly averages, aggregation to 168–173

MONYY5. format 9–10

MONYY7. format 9–10

moving averages 40–41, 124

N

NLDATE12. format 11

NOEST option

SEASON statement (UCM) 168

SLOPE statement (UCM) 143, 151–153, 162

NOINT option, MODEL statement (VARMAX) 84

nonseasonal exponential smoothing

about 29–30, 53

applying confidence limits for forecasts 42–43

basic formula 40–41

calculating confidence limits for forecasts 42

damped-trend method 49–53

Danish fertility example 33–40, 44

determining smoothing constant 46–47

double 30–33, 36, 42

estimating smoothing parameter in ESM procedure 47–49

Holt method 49–50

moving averages and 40–41

NOPRINT option, OUTPUT statement (VARMAX) 91

normal distribution, in exponential smoothing 42

NSEASON= option, MODEL statement (VARMAX) 94

O

OBS= option, PRINT procedure 12

ODS Graphics System 91, 104

ODS OUTPUT statement 120

OECD (Organization for Economic Co-operation and Development) 86

Organization for Economic Co-operation and Development (OECD) 86

OUT= option, OUTPUT statement (VARMAX) 91

OUTFOR= option, ESM procedure 33–34

OUTLIER statement

UCM procedure 137, 139

X12 procedure 121–122

outliers

additive 115, 117, 121–122, 139–140

Arctic ice coverage example 185–186

deleting 185

exponential smoothing of 43, 57, 61

UCM procedure and 139–142

X11 method and 103–104, 111, 113, 115

X12 procedure and 115–118, 121–122

output formats for datetime variables 9–12

OUTPUT statement

VARMAX procedure 82, 91

X12 procedure 104–108, 110–113, 118–120,
124–128

P

parameterized models versus exponential smoothing 75–95

percent sign (%) 9

period (.) 9

PLOT= option

ESM procedure 44

ESTIMATE statement (UCM) 139, 141, 146, 158, 161, 167

LEVEL statement (UCM) 138, 142

RANDOMREG statement (UCM) 157

SEASON statement (UCM) 154

SLOPE statement (UCM) 142

VARMAX procedure 83

PLOTS= option, UCM procedure 169, 182

prediction errors

Danish fertility example 34, 39–40, 52

Danish hotel overnight stays by US citizens example 61

e-commerce example 66, 69, 72, 74

exponential smoothing and 38–39

Swiss Business Indicator 89–90

PRINT= option

ESM procedure 33–34, 47, 51, 66, 87

LEVEL statement (UCM) 142

MODEL statement (VARMAX) 94

SEASON statement (UCM) 169

PRINT procedure

FIRSTOBS= option 12

OBS= option 12

procedure calls, including formats in 10–12

R

RANDOMREG statement, UCM procedure

about 156–157

PLOT= option 157

regression coefficient 155–158

REGRESSION statement, X12 procedure 116, 122

S

SAS/ETS software 5

SBC (Schwarz’s Bayesian Criterion) 77, 91

Schwarz’s Bayesian Criterion (SBC) 77, 91

SEASON statement, UCM procedure

about 149, 157, 174, 177

KEEPH= option 175

NOEST option 168

PLOT= option 154

PRINT= option 169

VAR= option 168

seasonal adjustments

unobserved components models and 133

X11 method for 99–113

X12 procedure and 115–128

seasonal exponential smoothing

about 55–56

additive smoothing 61–63

Danish hotel overnight stays by US citizens example 57–61

e-commerce examples 63–74

ESM procedure supporting 61–63, 67–74

Winters method 55–56, 63–68

seasonality 99–101

SET statement 120

SETMISSING= option, TIMESERIES procedure 21

SGPLOT procedure

Danish fertility example 33–34, 39–40

Danish hotel overnight stays by US citizens example 57–58, 124–128

e-commerce example 104–105, 118–120

time series aggregation example 20–21

time series interpolation example 24

SLOPE statement, UCM procedure

about 149, 157

NOEST option 143, 151–153, 162

PLOT= option 142

VAR= option 143, 151–153, 162

smoothing constant 46–47

smoothing parameter, ESM procedure 47–49

SORT procedure 16–17

state space models 135

STATESPACE procedure 135

Statistical Graphics facility 43–44

STEPwise AutoRegression (STEPAR) method 78, 90

Student’s t-distribution 39, 42

Swiss Business Indicator example

ESM procedure and 86–90

VARMAX procedure and 90–95

T

time formats in procedure calls 12

time series

additional information 6

aggregation of 19–22, 161–186

datetime variables 9–18

examples of 3–4

interpolation of 23–26

SAS procedures supporting 5–6

trends in 101–103, 124–125

types of 4–5

TIMESERIES procedure

about 5, 161

aggregation example 19–22, 161–186

ID statement 168, 179

SETMISSING= option 21

WHERE statement 161

TO= option, EXPAND procedure 23

TOD12. format 11

TRANSFORM= option, FORECAST statement (ESM) 74

TRANSFORM statement, X12 procedure 116

TRANSFORMOUT= option, CONVERT statement (EXPAND) 25

TREND= option, MODEL statement (VARMAX) 94

TRENDMA= option, X11 statement (X12) 124

trends in time series 101–103, 124–125

TYPE= option, OUTLIER statement (X12) 121

U

UCM procedure

See also ESTIMATE statement, UCM procedure

See also SEASON statement, UCM procedure

See also SLOPE statement, UCM procedure

about 5, 25

Arctic ice coverage example 161–185

AUTOREG statement 164

CYCLE statement 145, 176–177

Danish fertility example 137–148

DEPLAG statement 146–147, 172, 177

e-commerce example 108, 149–160

FORECAST statement 138, 163–164

ID statement 137, 161, 168, 181

IRREGULAR statement 172, 182

LEVEL statement 138, 142, 149, 157

MODEL statement 137, 156

OUTLIER statement 137, 139

PLOTS= option 169, 182

RANDOMREG statement 156–157

state space models and 135

TIMESERIES procedure and 161

unobserved components models and 131

unemployment in UK example 108–113

UNIVARIATE procedure 137

unobserved components models

about 131–132

Arctic ice coverage example 161–186

ARIMA representation 132

Danish fertility example 137–148

e-commerce example 149–160

estimation of 134–135

extensions of 132–134

state space models and 135

V

VAR= option

SEASON statement (UCM) 168

SLOPE statement (UCM) 143, 151–153, 162

VARMAX procedure

about 5–6

ARIMA (Box-Jenkins) models and 81–82

ARMA models and 172

Danish fertility example 82–85

fitted ARMA models and 82–85

ID statement 82

MODEL statement 82, 84, 91, 94

OUTPUT statement 82, 91

PLOT= option 83

Swiss Business Indicator example 90–95

X12 procedure and 116

W

weekly averages, aggregation to 173–179

WHERE= option, SET statement 120

WHERE statement, TIMESERIES procedure 161

Winters method

Danish hotel overnight stays by US citizens example 62–63

e-commerce examples 63–68

for seasonal forecasting 56–57

transformation and 70

X

X11-ARIMA method, X12 procedure 101, 116–117

X11 method

about 101–103

e-commerce example 103–108

“forgetting” algorithm 113

outliers and 115

unemployment in UK example 108–113

X11 procedure 101, 104

X11 statement, X12 procedure 123

MODE= option 121

TRENDMA= option 124

X12 procedure

about 5, 115–116

ARIMA (Box-Jenkins) models and 116

ARIMA statement 118

AUTOMDL statement 117–118, 123

X12 procedure (continued)

Danish hotel overnight stays by US citizens example 121–128

e-commerce example 104, 108, 116–120

model fitting and forecasting 115–116

OUTLIER statement 121–122

OUTPUT statement 104–108, 110–113, 118–120, 124–128

REGRESSION statement 116, 122

seasonal adjustment and 116–128

TRANSFORM statement 116

UCM procedure and 160

unemployment in UK example 110–113

X11-ARIMA method 101, 116–117

X11 statement 121, 123–124

Y

yearly averages, aggregation to 161–168

Symbols

. (period) 9

% (percent sign) 9

$ (dollar sign) 9

..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
3.145.75.217