Scribd
Upload
164 views11 pages

Eviews Brief Tutorial PDF

This document provides a brief overview of key functions in EViews for importing and preparing data, performing regressions and forecasting, and testing models and time series characteristics. Data can be imported via the foreign data wizard or by copying into a new series. New series can be generated using transformations. Regressions are estimated using the estimation equation wizard. Forecasting is done using the forecasting wizard on estimated models. Model goodness and residual diagnostics are assessed using coefficient, residual, and stability diagnostics views. Time series properties are tested using unit root and other tests accessible in the series spreadsheet.

Uploaded by

HarunVeledar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
164 views11 pages

Eviews Brief Tutorial PDF

This document provides a brief overview of key functions in EViews for importing and preparing data, performing regressions and forecasting, and testing models and time series characteristics. Data can be imported via the foreign data wizard or by copying into a new series. New series can be generated using transformations. Regressions are estimated using the estimation equation wizard. Forecasting is done using the forecasting wizard on estimated models. Model goodness and residual diagnostics are assessed using coefficient, residual, and stability diagnostics views. Time series properties are tested using unit root and other tests accessible in the series spreadsheet.

Uploaded by

HarunVeledar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 11

A Very Brief EViews

Tutorial
Contents
Importing data ................................................................................................................................................... 2
Transformations and generating new series ..................................................................................................... 4
Drawing graphs .................................................................................................................................................. 6
Regressions ........................................................................................................................................................ 7
Forecasting ........................................................................................................................................................ 9
Testing ............................................................................................................................................................. 10

1
Importing data
The easiest way to import data to Eviews is perhaps to use File -> Open -> Foreign data as Workfile

This starts a ready-made and easy-to-use wizard for importing data from .xls-files.

Data may also be copy pasted to an existing workfile. The drawback to this method is the fixed sample size
for the work file, i.e. it is tedious to import new data with sample size 1000 to a workfile that has existing
data with sample size 500.

In order to copy paste it is first necessary to generate a new blank series. This is done by using Object ->
New Object.

2
This generates a new series in the workfile window. To copy paste the data it is necessary to enable editing
by right clicking the series spread sheet:

After this the data may be copy pasted. If the copied data is larger than the workfile-specific sample size,
EViews automatically drops the ”excess” data.

3
Transformations and generating new series
Usually it is necessary to edit the imported raw data. For example, one may have to transform a price series
to a returns series. This can be done in EViews by using Quick -> Generate Series… :

This opens a box asking for equation:

Some of the more useful transformations are the logarithmic transformation: lnseries = log(series) and the
difference transformation: diffseries = d(series). The first generates a new series lnseries where each
observation is natural logarithm of the old series series, i.e. ( ). The second generates
a new series diffseries where each observation is the difference of two concurrent observations in series,
i.e. . Thus, it is possible to generate the returns series by the
transformation returns = d(log(prices)).

4
EViews also provides the possibility of generating residual series for regression models (running the actual
regressions is covered below). This may be used to gauge the goodness of the model. The residual series
can be generated from the equation box by using Proc -> Make Residual Series… :

5
Drawing graphs
The easiest way to draw graphs is to open the series spread sheet and use View -> Graph which generates a
graph-drawing wizard:

This can be used to draw the traditional graphs as well as the Q-Q plot for checking/testing distributions.

The menu in View-> Descriptive Statistics & Tests also provides useful graphs – notably Histogram and Stats
for analyzing distributions:

6
Regressions
The easiest way to run regressions in EViews is done using Quick -> Estimate Equation:

This provides a wizard for estimating equations:

The specification for the standard OLS-regression may be given by the formula:

dependent_variable c independent_variable1 independent_variable2 …

The wizard also recognizes some words as reserved, i.e. these cannot be used as independent variable
names. For example, the c in the formula refers to the constant term in the regression. Likewise, the ar(1)
and the ma(2) term refer to a first autoregressive term and a second moving average

7
term . It is also possible to lag terms by using the suffix (-i) for the ith lag, i.e. if we want to use the fifth
lag in the model we need to add the term diffexamplser(-5).

The wizard can also be used to use different methods. Similar specification guidelines should also hold for
the more exotic methods.

8
Forecasting
EViews also provides ready-made tools for forecasting. The easiest way to forecast is by using the
regression model windows Proc -> Forecast:

The forecasting wizard may need some extra explanations:

The wizard generates a new series that holds the forecasts, in this case the series diffexamplf. The method
should usually be set to static forecast. Likewise, the forecast sample should be set to appropriate
observations.

9
Testing
Typically we are interested in testing two issues – the goodness of a model and the characteristics of a time
series.

The goodness of model may be tested in the model window View -> Coefficient Diagnostics, Residual
Diagnostics and Stability diagnostics:

Residual diagnostics are especially useful when working with time series – the Correlogram – Q-statistics
gives information on the sufficiency of the model and Correlogram Squared Residuals gives information on
possible heteroscedasticity. Heteroscedasticity can and should also be tested with specific
Heteroscedasticity Tests.

The characteristics of a series can be tested in the series spread sheet:

In particular, Unit Root Tests for stationarity will be useful.

10

You might also like