NAG CL Interface
g13amc (uni_smooth_exp)
1
Purpose
g13amc performs exponential smoothing using either single exponential, double exponential or a Holt–Winters method.
2
Specification
void |
g13amc (Nag_InitialValues mode,
Nag_ExpSmoothType itype,
Integer p,
const double param[],
Integer n,
const double y[],
Integer k,
double init[],
Integer nf,
double fv[],
double fse[],
double yhat[],
double res[],
double *dv,
double *ad,
double r[],
NagError *fail) |
|
The function may be called by the names: g13amc, nag_tsa_uni_smooth_exp or nag_tsa_exp_smooth.
3
Description
Exponential smoothing is a relatively simple method of short term forecasting for a time series. g13amc provides five types of exponential smoothing; single exponential, Brown's double exponential, linear Holt (also called double exponential smoothing in some references), additive Holt–Winters and multiplicative Holt–Winters. The choice of smoothing method used depends on the characteristics of the time series. If the mean of the series is only slowly changing then single exponential smoothing may be suitable. If there is a trend in the time series, which itself may be slowly changing, then double exponential smoothing may be suitable. If there is a seasonal component to the time series, e.g., daily or monthly data, then one of the two Holt–Winters methods may be suitable.
For a time series
, for
, the five smoothing functions are defined by the following:
- Single Exponential Smoothing
- Brown Double Exponential Smoothing
- Linear Holt Smoothing
- Additive Holt–Winters Smoothing
- Multiplicative Holt–Winters Smoothing
and is defined as in the additive Holt–Winters smoothing,
where
is the mean,
is the trend and
is the seasonal component at time
with
being the seasonal order. The
-step ahead forecasts are given by
and their variances by
. The term
is estimated as the mean deviation.
The parameters, , and control the amount of smoothing. The nearer these parameters are to one, the greater the emphasis on the current data point. Generally these parameters take values in the range to . The linear Holt and two Holt–Winters smoothers include an additional parameter, , which acts as a trend dampener. For the trend is dampened and for the forecast function has an exponential trend, removes the trend term from the forecast function and does not dampen the trend.
For all methods, values for
,
,
and
can be chosen by trying different values and then visually comparing the results by plotting the fitted values along side the original data. Alternatively, for single exponential smoothing a suitable value for
can be obtained by fitting an
model (see
g13bec). For Brown's double exponential smoothing and linear Holt smoothing with no dampening, (i.e.,
), suitable values for
and
can be obtained by fitting an
model. Similarly, the linear Holt method, with
, can be expressed as an
model and the additive Holt–Winters, with no dampening, (
), can be expressed as a seasonal ARIMA model with order
of the form
. There is no similar procedure for obtaining parameter values for the multiplicative Holt–Winters method, or the additive Holt–Winters method with
. In these cases parameters could be selected by minimizing a measure of fit using one of the nonlinear optimization functions in
Chapter E04.
In addition to values for , , and , initial values, , and , for , are required to start the smoothing process. You can either supply these or they can be calculated by g13amc from the first observations. For single exponential smoothing the mean of the observations is used to estimate . For Brown double exponential smoothing and linear Holt smoothing, a simple linear regression is carried out with the series as the dependent variable and the sequence as the independent variable. The intercept is then used to estimate and the slope to estimate . In the case of the additive Holt–Winters method, the same regression is carried out, but a separate intercept is used for each of the seasonal groupings. The slope gives an estimate for and the mean of the intercepts is used as the estimate of . The seasonal parameters , for , are estimated as the intercepts – . A similar approach is adopted for the multiplicative Holt–Winter's method.
One step ahead forecasts,
are supplied along with the residuals computed as
. In addition, two measures of fit are provided. The mean absolute deviation,
and the square root of the mean deviation
4
References
Chatfield C (1980) The Analysis of Time Series Chapman and Hall
5
Arguments
-
1:
– Nag_InitialValues
Input
-
On entry: indicates if
g13amc is continuing from a previous call or, if not, how the initial values are computed.
- Required values for , and
, for , are supplied in init.
- g13amc continues from a previous call using values that are supplied in r.
- Required values for , and
, for , are estimated using the first observations.
Constraint:
, or .
-
2:
– Nag_ExpSmoothType
Input
-
On entry: the smoothing function.
- Single exponential.
- Brown double exponential.
- Linear Holt.
- Additive Holt–Winters.
- Multiplicative Holt–Winters.
Constraint:
, , , or .
-
3:
– Integer
Input
-
On entry: if
or
, the seasonal order,
, otherwise
p is not referenced.
Constraint:
if or , .
-
4:
– const double
Input
-
Note: the dimension,
dim, of the array
param
must be at least
- , when ;
- , when ;
- , when .
On entry: the smoothing parameters.
If
or
,
and any remaining elements of
param are not referenced.
If
,
,
,
and any remaining elements of
param are not referenced.
If or , , , and .
Constraints:
- if , ;
- if , ;
- if , and and ;
- if or , and and and .
-
5:
– Integer
Input
-
On entry: the number of observations in the series.
Constraint:
.
-
6:
– const double
Input
-
On entry: the time series.
-
7:
– Integer
Input
-
On entry: if
, the number of observations used to initialize the smoothing.
If
,
k is not referenced.
Constraints:
- if and or , ;
- if and , or , .
-
8:
– double
Input/Output
-
Note: the dimension,
dim, of the array
init
must be at least
- , when ;
- , when ;
- , when .
On entry: if
, the initial values for
,
and
, for
, used to initialize the smoothing.
If
,
and the remaining elements of
init are not referenced.
If
or
,
and
and the remaining elements of
init are not referenced.
If
or
,
,
and
to
hold the values for
, for
. The remaining elements of
init are not referenced.
On exit: if , the values used to initialize the smoothing. These are in the same order as described above.
-
9:
– Integer
Input
-
On entry: the number of forecasts required beyond the end of the series. Note, the one step ahead forecast is always produced.
Constraint:
.
-
10:
– double
Output
-
On exit:
, for
, the next
nf step forecasts. Where
, if
, else
is the total number of smoothed and forecast values already produced.
-
11:
– double
Output
-
On exit: the forecast standard errors for the values given in
fv.
-
12:
– double
Output
-
On exit: , for , the one step ahead forecast values, with being the one step ahead forecast of .
-
13:
– double
Output
-
On exit: the residuals,
, for .
-
14:
– double *
Output
-
On exit: the square root of the mean deviation.
-
15:
– double *
Output
-
On exit: the mean absolute deviation.
-
16:
– double
Input/Output
-
Note: the dimension,
dim, of the array
r
must be at least
- , when ;
- , when .
On entry: if
,
r must contain the values as returned by a previous call to
g05pmc or
g13amc,
r need not be set otherwise.
If
,
or
, only the first
elements of
r are referenced, otherwise the first
elements are referenced.
On exit: the information on the current state of the smoothing.
Constraint:
if
,
r must have been initialized by at least one previous call to
g05pmc or
g13amc with
, and
r should not have been changed since the last call to
g05pmc or
g13amc.
-
17:
– NagError *
Input/Output
-
The NAG error argument (see
Section 7 in the Introduction to the NAG Library CL Interface).
6
Error Indicators and Warnings
- NE_ALLOC_FAIL
-
Dynamic memory allocation failed.
See
Section 3.1.2 in the Introduction to the NAG Library CL Interface for further information.
- NE_BAD_PARAM
-
On entry, argument had an illegal value.
- NE_ENUM_INT
-
On entry, .
Constraint: if or , .
- NE_INT
-
On entry, .
Constraint: .
On entry, .
Constraint: .
- NE_INT_2
-
On entry, and .
Constraint: if and or , .
- NE_INT_3
-
On entry, , .
Constraint: if and or , .
- NE_INTERNAL_ERROR
-
An internal error has occurred in this function. Check the function call and any array sizes. If the call is correct then please contact
NAG for assistance.
See
Section 7.5 in the Introduction to the NAG Library CL Interface for further information.
- NE_MODEL_PARAMS
-
A multiplicative Holt–Winters model cannot be used with the supplied data.
- NE_NO_LICENCE
-
Your licence key may have expired or may not have been installed correctly.
See
Section 8 in the Introduction to the NAG Library CL Interface for further information.
- NE_REAL_ARRAY
-
On entry, .
Constraint: .
On entry, .
Constraint: if , .
On entry, .
Constraint: .
On entry, the array
r has not been initialized correctly.
7
Accuracy
Not applicable.
8
Parallelism and Performance
g13amc is not threaded in any implementation.
Single exponential, Brown's double exponential and linear Holt smoothing methods are stable, whereas the two Holt–Winters methods can be affected by poor initial values for the seasonal components.
See also the function document for
g05pmc.
10
Example
This example smooths a time series relating to the rate of the earth's rotation about its polar axis.
10.1
Program Text
10.2
Program Data
10.3
Program Results