problem with pasting graph into Word

[ege_man]

Dear Tom,
I was trying to past the rats rgf file into the attached word file, but as you see there are blank parts in the graph therefore graph does not cover the whole area. I have this problem only in my laptop not in my office computer. This problem may be because of high screen resolution of my laptop since I realized that the size of the graph is getting bigger when I decrease the screen resolution from 3200X1800 to 1920X1080. I don't know if anybody face the similar problem before.
Best Regards

[TomDoan]

So far as we know, this works fine with version 10.

[ege_man]

But I am also using rats version 10.

[TomDoan]

Save the graph to WMF on the computer with the high resolution and send it as an attachment to support@estima.com.

[ege_man]

I send it now.

[TomDoan]

That appears to be doing exactly what you asked it to. Was this generated inside an SPGRAPH with VFIELDS=4 and HFIELDS=2 perhaps?

Some programs will automatically clip a metafile down to the size of the actual content. Word doesn't.

[ege_man]

No I am using the following code to graph time-varying parameters. I also checked with version 8.1 get the same graph I send you.

Code: Select all

spgraph(vfields=2, hfields=1)
graph(footer=" (a) Time-varying regression coefficient of Oil Price", pattern,shading=raterise2, min=-0.4, max=0.4) 3
# b2 1997:02:03 *
# lower2 1997:02:03 * 2
# upper2 1997:02:03 * 2
graph(footer="(b) Time-varying regression coefficient of Exchange Rate",pattern,shading=raterise3, min=-1.25, max=1.0) 3
# b3 1997:02:03 *
# lower3 1997:02:03 * 2
# upper3 1997:02:03 * 2
spgraph(done)

[TomDoan]

Could you post the whole program?

[ege_man]

Dear Tom
First part of the code is below, as I said the problem is solved when I plot graphs with lower screen resolution such as the graph in the attached word file is obtained with 1024X768.

Code: Select all

open data multifactor_2018_tr_dec_6.xlsx
calendar(d) 1997:01:03
data(format=xlsx,org=columns) 1997:01:03 *
table
set stret = TKBASMT
linreg(robust) stret
# constant tknatal EIAEBRT2 turkli
*graph(key=below) 2
*# EIAEBRT2
*# EIAEBRT
equation(lastreg) mdeq
source rollcorr.src
@rollcorr(window=750,method=centered,results=oil_ret_rcor0,lags=0, nograph) stret EIAEBRT2
@rollcorr(window=750,method=centered,results=er_ret_rcor0,lags=0, nograph) stret TURKLI

stats(fractile) oil_ret_rcor0
compute varianceret_oil=%variance
compute stdevret_oil=sqrt(varianceret_oil)

stats(fractile) er_ret_rcor0
compute varianceret_er=%variance
compute stdevret_er=sqrt(varianceret_er)

set upper_oil  = oil_ret_rcor0+2.5*stdevret_oil
set lower_oil = oil_ret_rcor0-2.5*stdevret_oil
set upper_er  = er_ret_rcor0+2.5*stdevret_er
set lower_er = er_ret_rcor0-2.5*stdevret_er

set low_oil = upper_oil<0.and.lower_oil<0
set low_er = upper_er<0.and.lower_er<0
spgraph(vfields=2, hfields=1)
graph(header="Correlation with oil prices", shading=low_oil, pattern, min=-1.00, max=1.00) 3
# oil_ret_rcor0 1997:02:03 *
# upper_oil 1997:02:03 * 2
# lower_oil 1997:02:03 * 2
graph(header="Correlation with exchange rate",shading=low_er, pattern, min=-1.00, max=1.00) 3
# er_ret_rcor0 1997:02:03 *
# upper_er 1997:02:03 * 2
# lower_er 1997:02:03 * 2
spgraph(done)

[ege_man]

I also get a fully fitted graph with 1600X900 resolution.

[TomDoan]

Just to try to figure out what's happening with this, you're changing the resolution on a single computer? At what points do you change it?