Добавил:
Upload Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:
Eviews5 / EViews5 / Docs / EViews 5 Command Ref.pdf
Скачиваний:
75
Добавлен:
23.03.2015
Размер:
5.23 Mб
Скачать

512—Appendix B. Command Reference

trace

Model View

 

 

Display trace view of a model showing iteration history for selected solved variables.

Syntax

Model View:

model_name.trace(options)

Options

p

Print the block structure view.

 

 

Cross-references

See “Diagnostics” on page 811 of the User’s Guide for further details on tracing model solutions.

See also msg (p. 371), solve (p. 475) and solveopt (p. 476).

tramoseats

Series Proc

 

 

Run the external seasonal adjustment program Tramo/Seats using the data in the series.

tramoseats is available for annual, semi-annual, quarterly, and monthly series. The procedure requires at least n observations and can adjust up to 600 observations where:

 

36

for monthly data

n =

 

(B.2)

max{12, 4 s}

for other seasonal data

Syntax

Series Proc: series_name.tramoseats(options) [base_name]

Enter the name of the original series followed by a dot, the keyword, and optionally provide a base name (no more than 20 characters long) to name the saved series. The default base name is the original series name. The saved series will have postfixes appended to the base name.

 

 

tramoseats—513

 

 

 

 

Options

 

 

 

 

 

 

runtype=arg

Tramo/Seats Run Specification: “ts” (run Tramo fol-

 

 

(default=“ts”)

lowed by Seats; the “opt=” options are passed to

 

 

 

Tramo, and Seats is run with the input file returned

 

 

 

from Tramo), “t” (run only Tramo), “s” (run only

 

 

 

Seats).

 

 

 

 

 

 

save=arg

Specify series to save in workfile: you must use one or

 

 

more from the following key word list: “hat” (forecasts

 

 

of original series), “lin” (linearized series from Tramo),

 

 

“pol” (interpolated series from Tramo), “sa” (seasonally

 

 

adjusted series from Seats), “trd” (final trend compo-

 

 

nent from Seats), “ir” (final irregular component from

 

 

Seats), “sf” (final seasonal factor from Seats), “cyc”

 

 

(final cyclical component from Seats).

 

 

To save more than one series, separate the list of key

 

 

words with a space. Do not use commas within the list

 

 

of save series.

 

 

The special key word “save=*” will save all series in

 

 

the key word list. The five key words “sa”, “trd”, “ir”,

 

 

“sf”, “cyc” will be ignored if “runtype=t”.

 

 

 

 

opt=arg

A space delimited list of input namelist. Do not use

 

 

 

commas within the list. The syntax for the input namel-

 

 

 

ist is explained in the.PDF documentation file. See also

 

 

 

“Notes” on page 513 below.

 

 

 

 

 

 

reg=arg

A space delimited list for one line of reg namelist. Do

 

 

not use commas within the list. This option must be

 

 

used in pairs, either with another “reg=” option or

 

 

“regname=” option. The reg namelist is available only

 

 

for Tramo and its syntax is explained in the .PDF docu-

 

 

mentation file. See also “Notes” on page 513 below.

 

 

 

 

regname=arg

Name of a series or group in the current workfile that

 

 

 

contains the exogenous regressors specified in the pre-

 

 

 

vious “reg=” option. See “Notes” on page 513 below.

 

 

 

 

 

 

p

Print the results of the Tramo/Seats procedure.

Notes

 

 

The command line interface to Tramo/Seats does very little error checking of the command syntax. EViews simply passes on the options you provide “as is” to Tramo/Seats. If the

514—Appendix B. Command Reference

syntax contains an error, you will most likely to see the EViews error message “output file not found”. If you see this error message, check the input files produced by EViews for possible syntax errors as described in “Trouble Shooting” on page 346 of the User’s Guide.

Additionally, here are some of the more commonly encountered syntax errors.

To replicate the dialog options from the command line, use the following input options in the “opt=” list. See the .PDF documentation file for a description of each option.

1.data frequency: “mq=”.

2.forecast horizon: “npred=” for Tramo and “fh=” for Seats.

3.transformation: “lam=”.

4.ARIMA order search: “inic=” and “idif=”.

5.Easter adjustment: “ieast=”.

6.trading day adjustment: “itrad=”.

7.outlier detection: “iatip=” and “aio=”.

The command option input string list must be space delimited. Do not use commas. Options containing an equals sign should not contain any spaces around the equals; the space will be interpreted as a delimiter by Tramo/Seats.

If you set “rtype=ts”, you are responsible for supplying either “seats=1” or “seats=2” in the “opt=” option list. EViews will issue the error message “Seats.itr not found” if the option is omitted. Note that the dialog option Run Seats after Tramo sets “seats=2”.

Each “reg=” or “regname=” option is passed to the input file as a separate line in the order that they appear in the option argument list. Note that these options must come in pairs. A “reg=” option must be followed by another “reg=” option that specifies the outlier or intervention series or by a “regname=” option that provides the name for an exogenous series or group in the current workfile. See the sample programs in the “./Example Files” directory.

If you specify exogenous regressors with the “reg=” option, you must set the appropriate “ireg=” option (for the total number of exogenous series) in the “opt=” list.

To use the “regname=” option, the preceding “reg=” list must contain the “user=- 1” option and the appropriate “ilong=” option. Do not use “user=1” since EViews will always write data in a separate external file. The “ilong=” option must be at least the number of observations in the current workfile sample plus the number of forecasts. The exogenous series should not contain any missing values in this range.

Note that Tramo may increase the forecast horizon, in which case the exogenous series is extended by appending zeros at the end.

tsls—515

Examples

freeze(tab1) x.tramoseats(runtype=t, opt="lam=-1 iatip=1 aio=2 va=3.3 noadmiss=1 seats=2", save=*) x

replicates the example file EXAMPLE.1 in Tramo. The output file from Tramo is stored in a text object named tab1. This command returns three series named X_HAT, X_LIN, X_POL.

show x.TramoSeats(runtype=t, opt="NPRED=36 LAM=1 IREG=3 INTERP=2 IMEAN=0 P=1 Q=0 D=0", reg="ISEQ=1 DELTA=1.0", reg="61 1", reg="ISEQ=8 DELTAS=1.0", reg="138 5 150 5 162 5 174 5 186 5 198 5 210 5 222 5", reg="ISEQ=8 DELTAS=1.0", reg="143 7 155 7 167 7 179 7 191 7 203 7 215 7 227 7") x

replicates the example file EXAMPLE.2 in Tramo. This command produces an input file containing the lines:

$INPUT NPRED=36

LAM=1

IREG=3 INTERP=2 IMEAN=0

P=1 Q=0 D=0, $

$REG ISEQ=1

DELTA=1.0$

 

 

 

 

 

 

 

 

 

61

1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$REG ISEQ=8

DELTAS=1.0$

 

 

 

 

 

 

 

 

138

5

150

5

162

5

174

5

186

5

198

5

210

5

222

5

$REG ISEQ=8

DELTAS=1.0$

 

 

 

 

 

 

 

 

143

7

155

7

167

7

179

7

191

7

203

7

215

7

227

7

Additional examples replicating many of the example files provided by Tramo/Seats can be found in the “./Example Files” directory. You will also find files that compare seasonal adjustments from Census X12 and Tramo/Seats.

Cross-references

See “Tramo/Seats” on page 342 of the User’s Guide for discussion. See also the Tramo/ Seats documentation that accompanied your EViews distribution.

See seas (p. 440) and x12 (p. 550).

tsls

Command || Equation Method | Pool Method | System Method

 

 

Two-stage least squares.

Carries out estimation for equations, pools, or systems using two-stage least squares.

516—Appendix B. Command Reference

Syntax

Command:

tsls(options) y x1 [x2 x3 ...] @ z1 [z2 z3

...]

 

 

tsls(options) specification @ z1 [z2 z3 ...

]

 

Equation Method: eq_name.tsls(options) y x1

[x2 x3 ...

] @ z1 [z2 z3 ...]

 

eq_name.tsls(options) specification @ z1 [z2 z3

...]

Pool Method:

pool_name.tsls(options) y x1 [x2 x3 ...

] [@cxreg w1 w2 ...] [@per-

 

reg w3 w4 ...] [@inst z1 z2

...] [@cxinst z3 z4 ...

] [@perinst z5 z6

 

...]

 

 

 

 

System Method:

system_name.tsls(options)

 

 

 

 

To use tsls as a command or equation method, list the dependent variable first, followed by the regressors, then any AR or MA error specifications, then an “@”-sign, and finally, a list of exogenous instruments. You may estimate nonlinear equations or equations specified with formulas by first providing a specification, then listing the instrumental variables after an “@”-sign.

There must be at least as many instrumental variables as there are independent variables. All exogenous variables included in the regressor list should also be included in the instrument list. A constant is included in the list of instrumental variables even if not explicitly specified.

Options

General options

m=integer

Set maximum number of iterations.

 

 

c=number

Set convergence criterion. The criterion is based upon

 

the maximum of the percentage changes in the scaled

 

coefficients.

 

 

deriv=keyword

Set derivative methods. The argument keyword should

 

be a oneor two-letter string. The first letter should

 

either be “f” or “a” corresponding to fast or accurate

 

numeric derivatives (if used). The second letter should

 

be either “n” (always use numeric) or “a” (use analytic

 

if possible). If omitted, EViews will use the global

 

defaults.

 

 

showopts /

[Do / do not] display the starting coefficient values and

-showopts

estimation options in the estimation output.

 

 

p

Print estimation results.

 

 

tsls—517

Additional Options for Non-Panel Equation estimation

w=series_name

Weighted TSLS. Each observation will be weighted by

 

multiplying by the specified series.

 

 

h

White’s heteroskedasticity consistent standard errors.

 

 

n

Newey-West heteroskedasticity and autocorrelation

 

consistent (HAC) standard errors.

 

 

s

Use the current coefficient values in “C” as starting val-

 

ues for equations with AR or MA terms (see also param

 

(p. 404)).

 

 

s=number

Determine starting values for equations specified by list

 

with AR or MA terms. Specify a number between zero

 

and one representing the fraction of preliminary least

 

squares estimates computed without AR or MA terms.

 

Note that out of range values are set to “s=1”. Specify-

 

ing “s=0” initializes coefficients to zero. By default,

 

EViews uses “s=1”.

 

 

z

Turn off backcasting in ARMA models.

Additional Options for Pool and Panel Equation estimation

cx=arg

Cross-section effects. For fixed effects estimation, use

 

“cx=f”; for random effects estimation, use “cx=r”.

 

 

per=arg

Period effects. For fixed effects estimation, use “cx=f”;

 

for random effects estimation, use “cx=r”.

 

 

wgt=arg

GLS weighting: (default) none, cross-section system

 

weights (“wgt=cxsur”), period system weights

 

(“wgt=persur”), cross-section diagonal weighs

 

(“wgt=cxdiag”), period diagonal weights (“wgt=per-

 

diag”).

 

 

cov=arg

Coefficient covariance method: (default) ordinary,

 

White cross-section system robust (“cov=cxwhite”),

 

White period system robust (“cov=perwhite”), White

 

heteroskedasticity robust (“cov=stackedwhite”), Cross-

 

section system robust/PCSE (“cov=cxsur”), Period sys-

 

tem robust/PCSE (“cov=persur”), Cross-section het-

 

eroskedasticity robust/PCSE (“cov=cxdiag”), Period

 

heteroskedasticity robust (“cov=perdiag”).

518—Appendix B. Command Reference

keepwgts

Keep full set of GLS weights used in estimation with

 

object, if applicable (by default, only small memory

 

weights are saved).

 

 

rancalc=arg

Random component method: Swamy-Arora (“ran-

(default=“sa”)

calc=sa”), Wansbeek-Kapteyn (“rancalc=wk”), Wal-

 

lace-Hussain (“rancalc=wh”).

 

 

nodf

Do not perform degree of freedom corrections in com-

 

puting coefficient covariance matrix. The default is to

 

use degree of freedom corrections.

 

 

coef=arg

Specify the name of the coefficient vector (if specified

 

by list); the default is to use the “C” coefficient vector.

 

 

iter=arg

Iteration control for GLS specifications: perform one

(default=“onec”)

weight iteration, then iterate coefficients to convergence

 

(“iter=onec”), iterate weights and coefficients simulta-

 

neously to convergence (“iter=sim”), iterate weights

 

and coefficients sequentially to convergence

 

(“iter=seq”), perform one weight iteration, then one

 

coefficient step (“iter=oneb”).

 

Note that random effects models currently do not per-

 

mit weight iteration to convergence.

 

 

s

Use the current coefficient values in “C” as starting val-

 

ues for equations with AR or MA terms (see also param

 

(p. 404)).

 

 

s=number

Determine starting values for equations specified by list

 

with AR terms. Specify a number between zero and one

 

representing the fraction of preliminary least squares

 

estimates computed without AR terms. Note that out of

 

range values are set to “s=1”. Specifying “s=0” initial-

 

izes coefficients to zero. By default, EViews uses “s=1”.

 

 

Additional options for systems

i

Iterate on the weighting matrix and coefficient vector

 

simultaneously.

 

 

s

Iterate on the weighting matrix and coefficient vector

 

sequentially.

 

 

o (default)

Iterate only on the coefficient vector with one step of

 

the weighting matrix.

 

 

Соседние файлы в папке Docs