[More
EViews information here.]
I.
Point and Click
EViews
is like most of the software designed to run in Windows environment. It
utilizes most of the features in Windows such as drawdown menus and on-line
help. Like most of the Windows applications, things can be done in two ways:
selecting commands from the menu or typing in the commands directly. The latter
requires knowledge of the commands in EViews. The former is simply clicking the
items you want from the menu. This document is to show you how to do the former
in EViews. There are different ways of doing the same things. Feel free to
experiment.
Create
a Workfile
A
workfile MUST be created before doing anything else.
Click
File/New/Workfile
Select
"type of data" (annual? undated? ....)
Specify
"dates" (use help menu to learn the syntax of giving dates)
For
example,
|
|
Start date |
End date |
|
Annual |
1977 |
1986 |
|
Monthly |
1980:03 |
1983:12 |
|
Weekly |
2:10:51 |
3:17:94 |
|
Undated |
1 |
250 |
(If
it's undated data, specify a big number in "End date", say, 500, make
sure the number specified is bigger than the number of observations.)
Click
ok (A Workfile window pops up.)
*Different
windows will come up as you go; e.g., Workfile, Series, Group, etc. To
distinguish them, look at the title given at the top of each window.
Get
Data in EViews
Key
in data
Click
Objects/New Object ... at the top of the Econometric Views window
Select
Series from the "Type of object" list
Type
in the variable name, say, X, at the "Name of Object" slot
(An
object "X" is created in the Workfile window.)
Double
click the object X in the Workfile window
Click
Edit H/! at the top of the X window (to type in the values of X)
Click
on the first cell and type in the value of observation 1
Click
on the second cell and type in the value of observation 2, etc.
Import
data from Excel
Click
File/Import ... at the tope of the Econometric Views window
(Choose
Excel.xls from the "List Files of Type" at the bottom left hand
corner of the opened window.)
Select
filename
Click
ok
Specify
(1) "Order of data"
as
by observations if data are stored in the following form:
|
X |
Y |
|
1 |
3 |
|
4 |
5 |
|
6 |
7 |
|
5 |
9 |
or,
as by series if data are stored in the following form:
|
X |
1 |
4 |
6 |
5 |
|
Y |
3 |
5 |
7 |
9 |
(2)
"Upper-left data cell" (Give the location of the first data value.)
For
example, an Excel spreadsheet is given.
|
|
A |
B |
C |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Y
The "Upper-left data cell" is A2.
(3)
"Names for series or Number of series if names in file"
Type
in variable names with a space in between if variable names are not in the
Excel file. (For the above example, type in X Y)
If
the names are in file (e.g., above example), type in the number of series. (For
the above example, the number of series is 2)
(An
object for each variable is created in the Workfile window. Double click each
object to check on the data.)
Click
ok
Basic
Data Analysis - Summary Statistics, Graphs, and Regression Analysis
Suppose
two variable objects, X and Y, have been created in Workfile window.
-- Summary
statistics for one variable
Double
click X in the Workfile window (A window displaying all observations of
variable X appears.)
Click
View at the top of the X window and choose any of the statistics and
graphs you want to do. (check them out.)
-- Summary
statistics between two variables
Hold
down the left mouse button to select more than one object from the Workfile
window.
Double
click the selected objects
Click
Open Group (A Group window appears.)
Click
View and check out all the summary statistics and graphs you can get.
To
save the selected group of variables as an object in the Workfile window to use
it again later, do the following:
Click
Name at the top of the Group window (An Object Name window appears.)
Give
a name, say, GROUP1.
Click
ok (An object called GROUP1 is created in the Workfile window. Double
click the GROUP1 object to recall the Group window.)
To
save summary statistics for this group of variables, do the following.
Click
View/Descriptive Stats/Common Sample at the top of the Group window
Click
Freeze at the top of Group window (A Table window appears.)
(You
may give a title or edit the Table by clicking the items at the top of the
Table window.)
Click
Name (An Object Name window appears.)
Give
a name, say, TABLE1.
Click
ok (An object called TABLE1 is created in the Workfile window. Double
click the TABLE1 object to recall the regression results.)
*In
general, to save any result as an object in the workfile, click Freeze
and Name it!
-- Scatter
plot of two variables
Suppose
we want to plot Y against X.
Hold
down the left mouse button to select both Y and X objects from the Workfile
window. If the objects X and Y are not adjacent to each other, hold down the
Ctrl key and click on the objects X and Y. Make sure Y is selected FIRST.
Double
click the selected objects.
Click
Open Group
Click
View/Graph (A graph will appear.)
Double
click the graph to get to the Graph Options window.
Select
Scatter Diagram from the AGraph Type@ list and select other features you
want for your scatter plot.
Click
ok (A Group window appears with the scatter plot.)
To
save the plot as an object so it can be recalled later, do the following.
Click
Freeze at the top of Group window (A Graph window appears.)
(You
may give a title or edit the graph by clicking the items at the top of the
Graph window.)
Click
Name at the top of the Graph window (An Object Name window appears.)
Give
a name, say, GRAPH1.
Click
ok (An object called GRAPH1 is created in the Workfile window. Double
click the GRAPH1 object to recall the graph.)
-- Regression
analysis
Suppose
we want to regress Y on X.
Hold
down the left mouse button to select both Y and X objects. Make sure Y is
selected FIRST.
Double
click the selected objects.
Click
Open Equation ... (An Equation Specification window will appear.)
Click
ok (An untitled Equation window will appear and contains the results.)
To
save the equation as an object, do the following.
Click
Name at the tope of the Equation window (An Object Name window appears.)
Give
a name, say, EQ1.
Click
ok (An object called EQ1 is created in the Workfile window. Double click
the EQ1 object to recall the equation.)
To
save the results as an object, do the following.
Click
Freeze at the top of Equation window (A Table window appears.)
(You
may give a title or edit the Table by clicking the items at the top of the
Table window.)
Click
Name (An Object Name window appears.)
Give
a name, say, TABLE2.
Click
ok (An object called TABLE2 is created in the Workfile window. Double
click the TABLE2 object to recall the regression results.)
Saving
Results on Disk
Click
Save at the top of the Workfile window
Give
a filename with an extension .wf1 (For example, exercise.wf1)
To
recall the saved workfile, click File/Open at the Econometric
Views window and retrieve the file.
II.
Programming Commands
To
Write A Program:
Use
the mouse to click on FILE
Click
on NEW
At
ATYPE OF OBJECT@ prompt, click on PROGRAM
Click
on OK
Type
program statements
Click
on SAVE AS
Save
as PROGRAMNAME.PRG. Click OK
To
Run A Program
Create
a workfile first: click File-New-Workfile and specify a range of data (e.g., 1
200).
You
may add a line in your program to create the workfile. e.g., CREATE U 200 (see
help menu for options)
Type
RUN PROGRAMNAME on command line at top of screen or click on RUN on menubar in
the program window.
To
Plot Y against X
Click
on Y and X in the workfile window while pressing down the ctrl key.
Double
click the selected data and select Group. A window with two columns of numbers
appears.
Click
View-Graph. A graph window appears.
Double
click the graph. Another windows appears. Click Group Type and select Scatter
Diagram. Click ok.
To
save the graph as data, click Freeze and you may add title to the graph. Click
Name and give a name. An icon will be created in your workfile with the given
name. (You may "freeze" a table, a spreadsheet, or regression
results, and name them.) Double click on the icon to see the saved results.
When the workfile is saved, all data in the workfile, including the graph data,
will be saved. It can be retrieved next time by reopening the workfile.
To
Create A Table of Descriptive Statistics of Data in Workfile
Click
on the variable icons in workfile while holding down the ctrl key to select
variables for analysis.
Double
click the selected variables. Select AOpen Group@ and click View-Descriptive
Stats-Common Sample.
To
create a table icon of summary statistics in workfile, click Freeze and give a
name.
Tips
In
addition to using the pull-down menu, all commands can be typed in and executed
one by one at the command line which is the space below FILE.
Alternatively
you may first write a program file that contains all commands (e.g., create,
read, ls,...) to execute and run them at once.
In
a program file, use an apostrophe (>) before a comment statement if you do
not want Micro TSP to read it.
Save
a program file as filename.prg
Save
a workfile as filename.wf1
Use
the CREATE command at the beginning of a program to create a workfile.
example:
CREATE U 8 creates an undated (U) workfile with 8 observations.
CREATE
Q 46:1 88:1 creates a quarterly (Q) workfile from first quarter (.1)
1946 to first quarter 1988. CREATE A creates an annual workfile and CREATE
M creates a monthly workfile.
Use
the FETCH command to retrieve data stored as TSP data series.
example:
to retrieve the data series called GPA.dat and the data series called
INCOME.dat from the directory C:\EVIEWS\DATA use
FETCH
C:\EVIEWS\DATA\GPA C:\EVIEWS\DATA\INCOME
Use
the READ(o) command to retrieve data stored as an ASCII file. Click Help-Search
to search for on-line help on Read command for reading in different types of
data.
The
LS command returns the least squares linear regression equation.
example:
LS GPA C INCOME regresses the variable GPA on the variable INCOME
and a CONSTANT (C).
Use
the EQUATION command to name an equation.
example:
EQUATION EQ1.LS GPA C INCOME regresses GPA on INCOME and a
CONSTANT (C) and names the equation EQ1. A (group) object called EQ1
will be created in the workfile containing regression results once the above ls
command is executed.
example:
PRINT EQ1.STATS prints the results of the regression.
Alternatively,
the results of the regression can be printed without naming the equation.
example:
LS(P) GPA C INCOME prints(P) the results of the regression.
In
general, using (P) after a command sends the results to the printer and (S)
sends the results to the screen.
The
COEFCOV command generates the coefficient covariance matrix.
example:
PRINT EQ1.COEFCOV
The
RESIDS command generates a list of the residuals actual and fitted values.
example:
PRINT EQ1.RESIDS
In
general, using (G) after a command generates a graph.
example:
PRINT EQ1.RESIDS(G) prints a graph(G) of the residuals actual and
fitted values.
Use
the SMPL command to select a particular sample range.
example:
SMPL 59.1 88.1
Use
the GENR command to generate a new variable. Use the NRND to generate a
random number.
example:
GENR X=50+5*NRND creates a random variable called X.
Use
the R command to rename variables.
example:
R GCDAN S changes the name of the variable AGCDAN@ to AS@.
Use
(-1) or (-2) etc. after a variable to represent a lagged time period.
example:
GENR MAP=PDP+PDP(-1)+PDP(-2) generates a new variable called MAP which
is the three month sum of PDP.
Use
the SAVE command to save a workfile.
example:
SAVE INTEREST saves the workfile INTEREST for future use. When a
workfile is named and saved, everything within the workfile can be retrieved by
reopening the workfile.
Use
the LOG command to calculate the logarithm of a variable.
example:
GENR LPCEXP=LOG(PCEXP) generates a new variable LPCEXP which is
the logarithm of PCEXP.
Use
the @SSR, @SUM, and @COR commands to calculate the sum of squared residuals,
sum, and simple correlation respectively.
Use
the SCAT command to generate a scatter plot.
example:
SCAT(P) RES INC prints(P) a scatter plot of the dependent
variable RES versus the independent variable INC.
Use
the LS(W= ) command to perform a weighted least squares regression.
example:
LS(W=X) HEXP C INC performs a weighted regression with weight X.
Use
the PLOT command to create a graph.
example:
PLOT (A,P) E generates a graph of E and scales automatically (A)
and prints (P).
Use
the LOAD command to load a previously created workfile.
example:
LOAD INTEREST loads the workfile named INTEREST.
Use
the EXPAND command to extend the range of a workfile in order to forecast out
of sample.
example:
CREATE U 8 creates and undated workfile with 8 observations.
EXPAND
15 expands the workfile to 15 observations.
Use
the FIT command to produce a forecast.
Use
the SHOW command to see results on the screen.
example:
SHOW INCOME GPA shows the variables INCOME and GPA on the
screen.
Alternatively,
use the PRINT command to send results to the printer.
example:
PRINT INCOME GPA
Use
the IDENT command to produce the autocorrelation function and the partial
autocorrelation function.
example:
IDENT(#) where # is the number of correlations.
The
FORCST command computes an n-period ahead dynamic forecast for the dependent
variable in the active equation.
The
FIT command computes the fitted values (one period ahead forecast values) from
the active equation .
The
DIFFERENCE OPERATOR specifies the first difference of a variable (use D(GNP)
rather than @DIF(GNP)).
example:
D(GNP) computes GDP - GDP(-1).