RATS Dialog Interface
WINDOWS:
It is most convenient to use RATS using two windows: one for input and
the other for output. When you begin RATS, the ACTIVE, OUTPUT, and
INPUT are identical. This Window is NONAME00.TXT
If you open a second window without closing NONAME.TXT the new
window will become the ACTIVE window, but not the input or output
window. If yow want to use this new window for input and/or output, use
the Window-Use for Input or the Window-Use for Output commands listed
on the WINDOW menu.
The Window-Tile command, makes the multiple windows visible.
You can switch between windows using the mouse, menu or using <CTRL
F6>
Preparing the Data
1
RATS Dialog Interface: II
LOCAL MODE: <Ctrl L>
Allows you to enter commands or instructions, without execution taking
place. Thus, you can use RATS as a wordprocessor.
This is a toggle: the second <Ctrl L> returns you to ready mode.
SELECT TEXT: CUT, COPY, PASTE:
Use the mouse or use the SHIFT key in conjunction with the arrow keys or
the PgUp or PgDn keys to mark text. Marked text can be cut, pasted, or
copied using the EDIT menu. Note that <CTRL X> is equivalent to Cut,
<CTRL C> is equivalent to Copy, and <CTRL V> is equivalent to Paste.
COMPOUND INSTRUCTIONS:
Certain instructions have supplementary cards (e.g., the GRAPH
Command). If you make an error on a supplementary card, you must reenter
the entire instruction from the beginning.
Preparing the Data
2
RATS Dialog Interface: III
ONLINE HELP:
You can get online help from the HELP Pull-Down Menu. You can get help
on the specific RATS Instructions using the F1 key.
PRINTING GRAPHS:
To print a graph, you use the PRINT command from the FILE menu. RATS
will print only the material in the currently shown window. As such, the
graph window must be shown on the screen.
SAVING YOUR WORK:
It is possible to save your program on to a disk using the SAVE or the SAVE
AS command on the FILE menu. As such, you will not have to retype your
commands the next time you use RATS. Simply use the OPEN command
from the FILE menu to begin your next session. (Be sure the new program is
in the active window!).
Preparing the Data
3
Preparing the Data
The file labeled MONEY_DEM.XLS contains quarterly
values of seasonally adjusted U.S. nominal GDP, real GDP
in 1996 dollars (RGDP), the money supply as measured by
M2 and M3, and the 3-month and 1-year treasury bill rates
for the period 1959:1 – 2001:1. Both interest rates are
expressed as annual rates and the other variables are in
billions of dollars
cal 1959 1 4
all 2001:1
open data c:\money_dem.xls
data(org=obs,format=xls)
Preparing the Data
4
Calendar
calendar year period frequency
or
cal(q) 1959:1
where: year
period
frequency
The year of the first entry in the data set
The period of the first entry in the data set
The number of observations per year.
Examples:
For monthly data beginning with February 1973, use calendar 1973 2 12.
For semiannual data beginning with July 1973, use calendar 1973 2 2.
For annual data it is permissible to use only the starting year. As such, you
can omit the period and frequency and use the more compact calendar 1973
instead of calendar 1973 1 1.
Preparing the Data
5
Calendar II
CALENDAR(DAILY)
year month day
Sets date of first entry for business-day (5-day per week) daily data.
CALENDAR(PERDAY=# of periods) year month day
Sets starting date, frequency, and the number of periods per day for intraday
data
calendar(a) year:1
calendar(q) year:quarter
calendar(b) year:month:day
calendar(w) year:month:day
calendar(d) year:month:day
calendar(7) year:month:day
Annual
Quarterly
Biweekly
Weekly
Daily (5/week)
Daily (7/week)
Preparing the Data
6
Allocate
• In older versions, it was necessary to use
ALLOCATE immediately following
CALENDAR. However, in new versions it
is convenient, but not necessary.
• Examples:
– For monthly data ending with April 1990 use
allocate 1990:4.
For monthly data ending with December 1990, use
allocate 1990:12.
Preparing the Data
7
Open and DATA
The OPEN and DATA statements are used together.
data(format=xls,org=obs) start end series
where: start end Range of entries to read.
series
The list of series to read.
(If series is omitted, all variables in the data set are read into memory)
Notice the slash (/) in the DATA instruction. In RATS, you can set the range
explicitly or use a slash to refer to the default range.
Example:
cal(q) 1960:1
all 2012:4
open data quarterly(2012).xls
data(org=obs,format=xls)
Preparing the Data
8
Form of the Data Sets
DATE Tb3mo Tb1yr RGDP Potent Deflator
1960Q1 3.87 4.57 2845.3 2824.2 18.521
1960Q2 2.99 3.87 2832.0 2851.2 18.579
1960Q3 2.36 3.07 2836.6 2878.7 18.648
1960Q4 2.31 2.99 2800.2 2906.7 18.700
1961Q1 2.35 2.87 2816.9 2934.8 18.743
1961Q2 2.30 2.94 2869.6 2962.9 18.785
1961Q3 2.30 3.01 2915.9 2991.3 18.843
1961Q4 2.46 3.10 2975.3 3019.9 18.908
Preparing the Data
M2
298.7
301.1
306.5
310.9
316.3
322.1
327.6
333.3
PPI Curr
33.2 31.8
33.4 31.9
33.4 32.2
33.7 32.6
33.6 32.1
33.3 32.1
33.3 32.7
33.4 33.4
9
Sample Statistics
•
•
•
•
•
Table
table(picture="*.##")
statistics Tb3mo
table / tb3mo tb1yr
print / tb3mo tb1yr
Preparing the Data
10
SET
set series start end = function(T)
1. A slash (/) or the omission of start end both default to the maximum
permissible range as indicated on ALLOCATE.
2. If you make a mistake, RATS will create missing values instead of an
error message whenever possible.
THIS CAN BE A PROBLEM IF YOU ARE UNAWARE OF
THE MISTAKE!
3. You must use a space before and after the equal sign.
4. RATS accepts addition (+), subtraction (-), multiplication (*),
division (/), and exponentiation (**) using the usual order of
precedence.
5. ABS(X), EXP(X), LOG(X), and SQRT(X) are used to denote the
absolute value, exponential value (i.e., ex), natural logarithm, and
square root of the argument X, respectively.
6. All levels of parentheses ( ) are supported but you cannot use braces
{ } or brackets [ ] in place of parentheses.
Preparing the Data
11
SET: II
• RATS allows you to use braces { } as a shorthand way to represent
time subscripts; you can use series{lag} in place of series(t-lag).
Examples:
– Instead of using y(t-1), you can use y{1}. The number placed in
the braces { } indicates the lag number.
– A negative number in braces indicates a leading value of the
variable so that y(t+2) and y{-2} are equivalent ways to write the
second lead of y.
– Also, the notation {A to B} can be used to indicate lags A through
B so that y{1 to 4} indicates the first four lags of ja.
Preparing the Data
12
Set: III
•
•
•
•
•
•
set dlrgdp = log(rgdp) - log(rgdp{1})
set dlm2 = log(m2) - log(m2{1})
set drs = tb3mo - tb3mo{1}
set dr1 = tb1yr - tb1yr{1}
set dlp = log(deflator) - log(deflator{1})
set dlppi = log(ppi) - log(ppi{1})
Preparing the Data
13
Set: IV
• Shortcuts
– log y / ly
– difference y / y
• You must define the start end range or use the slash (/) on the
DIFFERENCE statement. Unlike the SET statement, if you
explicitly define the range, you must allow for the number of
lags created.
– exp y / ey
– seasonal seasons
– set trend = t
Preparing the Data
14
Graph
GRAPH(Options) number
# series start end symbol choice (one for each series)
number: number of series to graph
Options
STYLE=[LINE]/POLYGON/BAR/STACKEDBAR/OVERLAPBAR/
VERTICAL/STEP/SYMBOL
PATTERNS/[NOPATTERNS]
HEADER='string' SUBHEADER='label‘ HLABEL='label' VLABEL='label‘
NODATES (RATS will label the horizontal axis unless the NODATES option is
specified.)
KEY=[NONE]/UPLEFT/UPRIGHT/LOLEFT/LORIGHT/ABOVE/BELOW/
LEFT/RIGHT
OVERLAY=[LINE]/POLYGON/BAR/STACKEDBAR/OVERLAPBAR
/VERTICAL/STEP/SYMBOL
OVCOUNT=Number of series for right-side scale [1]
Preparing the Data
15
Examples
graph(header="The Interest Rates",key=below) 2
# tb3mo ; # tb1yr
graph(hea="Real and Potential GDP",key=upleft) 2
# rgdp ; # potent
Preparing the Data
16
Multiple Graphs per Page
SPGRAPH(hfields= ,vfields= ,header = ‘ ‘ , subheader = ‘ ‘) :
Spgraph(Done)
Example:
SPGRAPH(HFIELDS=2,VFIELDS=2)
Graph 1
# x1
….
Graph 1
# x4
SPGRAPH(DONE)
Preparing the Data
17
spgraph(footer="Graphs of the Series",hfields=2,vfields=2)
graph(header="Panel 1: The Interest Rates",key=below,nokbox) 2
# tb3mo ; # tb1yr
graph(header="Panel 2: Real and Potential GDP",key=upleft) 2
# rgdp ; # potent
graph(header="Panel 3: Time path of money growth",noaxis) 1
# dlm2
graph(header="Panel 4: Time path of Inflation",noaxis) 1
# dlp
spgraph(done)
Preparing the Data
18
Scatter Diagrams
SCATTER(Options) pairs
# x-series y-series start end
The important options are the same as those for Graph Instruction
NOTE:
ON COMPOUND STATEMENTS LIKE GRAPH AND SCATTER, IT
YOU MAKE A MISTAKE ON THE SECOND LINE, YOU MUST REENTER THE FIRST LINE.
Preparing the Data
19
Linear Regression
Suppose you want to estimate the regression
7
drst   0    i drst 1   t
i 1
lin drs / resids
# constant drs{1 to 7}
Preparing the Data
20
LINREG
linreg(options) depvar start end residuals
# list
where: depvar
start end
residuals
list
The dependent variable.
The range to use in the regression.
(Optional) Series name for the residuals.
The list of explanatory variables.
Preparing the Data
21
Linear Regression - Estimation by Least Squares
Dependent Variable DRS
Quarterly Data From 1962:01 To 2012:04
Usable Observations 204
Degrees of Freedom 196
Centered Rˆ2 0.2841953
R-Barˆ2 0.2586309
Uncentered Rˆ2 0.2843637
Mean of Dependent Variable -0.011617647
Std Error of Dependent Variable 0.759163288
Standard Error of Estimate 0.653660810
Sum of Squared Residuals 83.745401006
Regression F(7,196) 11.1168
Significance Level of F 0.0000000
Log Likelihood -198.6489
Durbin-Watson Statistic 1.9709
Variable Coeff Std Error T-Stat Signif
***********************************************************************************
*
1. Constant -0.011903358 0.045799634 -0.25990 0.79521316
2. DRS{1}
0.390010248 0.069644459 5.60002 0.00000007
3. DRS{2}
-0.380186642 0.074718282 -5.08827 0.00000084
4. DRS{3}
0.406843358 0.078304236 5.19567 0.00000051
5. DRS{4}
-0.159123423 0.082740231 -1.92317 0.05590809
6. DRS{5}
0.193334248 0.078290297 2.46945 0.01438724
7. DRS{6}
-0.089946745 0.074692035 -1.20423 0.22995107
8. DRS{7}
-0.220768119 0.069358921 -3.18298 0.00169542
Preparing the Data
22
LINREG: II
The most useful options are:
DEFINE = Name the equation by setting DEFINE = ‘NAME
NOPRINT
Do not print the regression output.
VCV Print the covariance/correlation matrix of the coefficients.
LINREG creates a number of variables that you can use in subsequent
computations. A partial list of these variables is:
%BETA
The coefficient vector. The first coefficient estimated is
%BETA(1), the second %BETA(2), and so on.
%tstats
%NDF
Degrees of freedom
%NOBS
Number of observations.
%NREG
Number of regressors.
%RSS
Residual sum of squares.
%RSQUARED Centered R2 (i.e, the usual measure of R2)
%SEESQ
Standard error of estimate squared
Preparing the Data
23
Correlations
CORRELATE: Calculates the autocorrelations (and the partial
autocorrelations) of a specified series. The syntax and principal options are:
correlate(options) series start end
where: series
The series used to compute the correlations.
start endThe range of entries to use. The default is the entire series.
The principal options are:
NUMBER=
The number of autocorrelations to compute.
PARTIAL=
Series for the partial autocorrelations.
QSTATSUse this option if you want the Ljung-Box Q-statistics.
DFC= Degrees of freedom correction.
SPAN= Use with qstats to set the width of the intervals tested. For example,
with quarterly data, you can set span = 4, to obtain Q(4), Q(8), Q(12), and so
forth.
Preparing the Data
24
correlate(options) series start end
where: series
The series used to compute the correlations.
start end
The range of entries to use. The default is the entire series.
corrs
Series used to save the autocorrelations (Optional).
The principal options are:
NUMBER= The number of autocorrelations to compute. The default is the integer value of one-fourth the total
number of observations.
PARTIAL= Series for the partial autocorrelations. If you omit this option, the PACF will not be calculated.
QSTATS
Use this option if you want the Ljung-Box Q-statistics.
SPAN=
Use with qstats to set the width of the intervals tested. For example, with quarterly data, you can set
span = 4, to obtain Q(4), Q(8), Q(12), and so forth.
In the example at hand, we can obtain the first twelve autocorrelations, partial autocorrelations (and the
associated Q-statistics) of the residuals with:
cor(number=12,partial=partial,qstats,span=4) resids
Preparing the Data
25
Restrictions
1. Exclusion and Sum Restrictions
Examples:
exclude
# drs{5 to 7}
exclude
# constant drs{5 to 7}
summarize
# drs{5 to 7}
Preparing the Data
26
TEST: Individual Coefficient Restrictions
2. Test
This instruction contains three lines:
TEST
# list of coefficients to be restricted
# constants against which coeff. are tested
Example:
To test if beta(6) = 0.1, beta(7) = 0.1 and %beta(8) = 0., use:
test
#678
# 0.1 0.1 0.0
To test the restriction that the constant equals zero and that 5 = 0.1, 6 = 0.1, and
7 = 0.1, use:
test
#1678
# 0. 0.1 0.1 0.1
Preparing the Data
27
Restrict
3. RESTRICT can test multiple linear restrictions on the
coefficients and estimate the restricted model. Each
restriction is entered in the form:
βiai + βjaj + ... + βkak = r
where: ai are the coefficients of the estimated model (i.e.,
each coefficient is referred to by its assigned number).
βi are weights you assign to each coefficient.
r represents the restricted value of the sum
Preparing the Data
28
Restrict: II
Examples
1. To test the joint restriction that a2 + 2a3 = 0 (i.e., 1.0*a2 + 2.0*a3 = 0), use:
restrict 1
#23
#120
2. To test the joint restriction that a2 + 2a3 = 0 (i.e., 1.0*a2 + 2.0*a3 = 0), use:
restrict 1
#23
#120
3. restrict(create) 2 resids
#5678
#11110
#23
# 1 -1 0
Preparing the Data
29
Forecast
FORECAST: Creates dynamic forecasts of a previously defined equation.
FORECAST uses a supplementary card for each equation in the system. In
univariate forecasting, the typical syntax for FORECAST is:
forecast(print) number steps start
# equation forecasts
where: number The number of equations in the system. In univariate
forecasting, number is necessarily equal to 1.
steps
The number of forecasts to create.
start
The starting period of the forecasts.
equationThe name of the previously defined equation.
forecastsThe name of the series in which you want to save the
forecasts. This field is optional.
Preparing the Data
30
Forecast: II
Example:
lin(define=eq1) rcons 1947:1 1997:2 resids
# constant trend t3 t4 t5
forecast(print) 1 12 1997:3
# eq1 fore
Preparing the Data
31