*
* Example from chapter 6, pp 241-263
*
open data bankdata.dat
calendar(m) 1988:2
data(format=free,org=columns) 1988:2 1993:1 eom aaa govt34
*
graph(footer="Figure 6-1 EOM Balances at a Mutual Savings Bank")
# eom
*
* On page 243, the differencing of the EOM and 3-4 rates are described as
* "forward" differences, that is, the value for period T is X(T+1)-X(T), rather
* than the more standard backward difference (X(T)-X(T-1)). To get X(T+1) in a
* SET instruction, use X{-1}.
*
set deom = eom{-1}-eom
set d34  = govt34{-1}-govt34
*
* The SPGRAPH(VFIELDS=4)...SPGRAPH(DONE) wrapper around the four GRAPH
* instructions is used to stack the four graphs onto a single page
*
spgraph(vfields=4)
graph(vlabel="D(EOM)")
# deom
graph(vlabel="AAA")
# aaa
graph(vlabel="(3-4)")
# govt34
graph(vlabel="D(3-4)")
# d34
spgraph(done)
*
@splom(labels=||"D(EOM)","AAA","(3-4)","D(3-4)"||)
# deom aaa govt34 d34
*
linreg deom * 53
# constant aaa govt34 d34
*
* The regression F-test is included in the output from LINREG. You can use the
* procedure REGANOVA to get the analysis of variance table as shown on page 254
*
@reganova
*
* If you want to see the covariance/correlation matrix of a regression, include
* the option VCV. Note that this shows the actual variances and covariances below
* the diagonal, and the correlations above it.
*
linreg(vcv) deom * 53
# constant aaa govt34 d34
*
prj fitted
*
spgraph(vfields=2,hfields=2,footer="Figure 6-4 Bank data: Plots of residuals")
scatter(vlabel="Residuals",hlabel="AAA",axis=horiz)
# aaa %resids
scatter(vlabel="Residuals",hlabel="D(3-4)",axis=horiz)
# d34 %resids
scatter(vlabel="Residuals",hlabel="(3-4)",axis=horiz)
# govt34 %resids
scatter(vlabel="Residuals",hlabel="Fitted",axis=horiz)
# fitted %resids
spgraph(done)
*
@histogram(maxgrid=12,distrib=normal,footer="Figure 6-5 Bank data: histogram of residuals") %resids