Graph adjustment

[Kaine Dinh]

Good afternoon everyone,

I am currently having an issue with the graph adjustment in Stata.
I am trying to replicate the provided graph as follow:

And my graph from Stata is displayed as below:

As you can see, the values of my vertical axis are the same as the model, but their position and scale is not the same.
How can I adjust them? The command that I used as follow:
Command: twoway (line residual Year) (line fitted_log_P Year, yaxis(2)) (line log_P Year, yaxis(2)), ylabel(-0.3(0.1)0.3) ylabel(-0.4(0.4)1.6, axis(2)) xlabel(1955(5)2005) xmtick(##5)
I am looking for your help. Best wishes!

[Andrew Musau]

Post the data. See FAQ Advice #12 for details on how to.

[Nick Cox]

I agree with Andrew Musau that a data example would help.

Meanwhile, an easy general comment is that these graphs displease somebody whichever way you do it.

If you plot residuals on the same scale as observed and fitted, that can be confusing -- with problems variously

a. the residual series being too flat to see any pattern

b. the residuals being so far from observed and fitted that too much white space appears

c. the residuals overlapping with observed and fitted, so three lines to disentangle.

If you plot residuals on one hand and observed and fitted on the other against different left- and right-hand scales

d. separate scales are more work for everyone to set up and decode

e. even linear scalings such as plotting residuals dilated and translated can be hard to read -- or at worst give rise to suspicion of sharp practice

f. in principle, all the information is there, but a Gestalt awareness of observed, fitted and residual can be elusive

Back in 2004, I published an ofrtplot that seems now on no-one's radar. The ugly name is "observed, fitted and residual versus t(ime) plot" where time is the usual suspect, but not the only one.

Here are the links

. search ofrtplot, sj

Search of official help files, FAQs, Examples, and Stata Journals

SJ-10-1 gr0009_1  . . . .  Software update for model diagnostic graph commands
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
        (help anovaplot, indexplot, modeldiag, ofrtplot, ovfplot,
        qfrplot, racplot, rdplot, regplot, rhetplot, rvfplot2,
        rvlrplot, rvpplot2 if installed)
        Q1/10   SJ 10(1):164
        provides new command rbinplot for plotting means or medians
        of residuals by bins; provides new options for smoothing
        using restricted cubic splines; updates anova examples

SJ-4-4  gr0009  . . . . . . . . . . Speaking Stata: Graphing model diagnostics
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
        (help anovaplot, indexplot, modeldiag, ofrtplot, ovfplot,
        qfrplot, racplot, rdplot, regplot, rhetplot, rvfplot2,
        rvlrplot, rvpplot2 if installed)
        Q4/04   SJ 4(4):449--475
        plotting diagnostic information calculated from residuals
        and fitted values from regression models with continuous
        responses

Here's a toy example: The graph shown is from the second command. The code is public if anyone wants something a bit or a lot different.

webuse grunfeld, clear

gen ln_invest = ln(invest)
regress ln_invest L.ln_invest year if company == 1 

set scheme s1color 
ofrtplot 
ofrtplot, rplot(recast(line) yli(0, lc(gs8)))