Plotting regression line for for different levels of a modertaing variable in a log-regression with interaction term

[Jakob Abekhon]

I have a simple bivariate regression model with an interaction term:

reg DP ln(IV) MOD ln(IV)*MOD

I want to plot the regression line for different values of MOD, to look something like the below:

I have tried to do this with margin and marginsplot commands, but I only get linear results (see attachment)

reg povertyline logdeficit gini interact_gini
margins, at(gini=(30 36 42))
marginsplot

I would be grateful for your help!

Kind regards,
Jakob

* Example generated by -dataex-. For more info, type help dataex
clear
input float iso3n double gini float(logdeficit interact_gini)
926                 29          .          .
499               36.9          .          .
 36                 34   3.330491   113.2367
554                  .   2.964826          .
 52                  .          .          .
136                  .          .          .
214               42.2  2.3475068   99.06479
780                  .  3.2702994          .
662                  .          .          .
308                  .          .          .
212                  .          .          .
670                  .          .          .
659                  .          .          .
533                  .          .          .
332                  .          .          .
 28                  .          .          .
850                  .          .          .
192                  .          .          .
630                  .          .          .
388                  .   2.432784          .
 44                  .          .          .
591               49.9   2.499323  124.71622
222                 38   1.840707   69.94687
340               49.4  1.6871662   83.34601
188               48.3  2.2543366  108.88446
320                  .   1.407686          .
558                  .  1.7681222          .
484 47.200000000000045  2.5772305  121.64528
 84                  .          .          .
860                  .          .          .
398               27.5   3.038216   83.55094
795                  .          .          .
762                  .   2.036664          .
417               27.3   2.340101   63.88476
450  43.09999999999985   .9080942   39.13886
834  40.11428571428564   .7790053   31.24924
262               41.6          .          .
716               44.3  1.8018782    79.8232
480               36.8    2.52604   92.95828
706                  .          .          .
231  35.72000000000003   1.251808   44.71458
232                  .          .          .
646  43.23333333333346   .6104927  26.393635
800 42.766666666666666  -.3747015 -16.024733
894  57.70000000000016  1.0381508    59.9013
174                  .          .          .
108                  .          .          .
690  35.04000000000087          .          .
508                  .  1.3518612          .
404                  .   1.290814          .
454  42.63333333333321 -2.3025696  -98.16622
446                  .          .          .
408                  .          .          .
156               39.1  2.7687166  108.25682
410  31.50000000000003  3.1917524   100.5402
496               32.5   2.960529   96.21719
344                  .   3.112752          .
392  33.96666666666624  3.1168535  105.86913
203               24.9  3.1798506   79.17828
804                 26   2.725242   70.85629
642                 36   2.639316   95.01538
498               25.9          .          .
100               40.4  2.7824554   112.4112
348               30.6  2.8531375   87.30601
112               25.4   2.814378    71.4852
810               37.2   3.039144  113.05617
616               29.7   2.986573   88.70123
703               23.2   2.929394   67.96194
598                  .          .          .
242  32.69999999999982          .          .
 90                  .          .          .
548 33.433333333333394          .          .
296                  .          .          .
583                  .          .          .
520                  .          .          .
316                  .          .          .
585                  .          .          .
584                  .          .          .
180                  .          .          .
266                 38          .          .
178                  .          .          .
148  38.32857142857142          .          .
226                  .          .          .
 24                  .          .          .
678               40.7          .          .
120                  .  1.4908074          .
140                  .          .          .
788  31.59999999999968  2.2922888  72.436325
504                  .   1.976289          .
434                  .          .          .
728                  .          .          .
736                  .          .          .
818               31.5   1.990596   62.70378
 12                  .          .          .
124               33.3   3.306146  110.09466
840               41.2   3.423671  141.05525
304                  .          .          .
 60                  .          .          .
578                 27   3.093841   83.53371
208               28.7   3.084956   88.53824
end

[Maarten Buis]

If you want to use margins you have to use the factor variable notation for the interaction variable and not create it yourself. See help fvvarlist

[Jakob Abekhon]

Thank you for your answer!

However, when I put in the following code

 reg povertyline logdeficit gini c.logdeficit#c.gini
margins logdeficit gini c.logdeficit#c.gini , at(gini=(30  36 42))

I receive the error message:
only factor variables and their interactions are allowed
r(198);

I'm sorry if these are very simple problems. I am a complete beginner.

Thank you for your help!

[George Ford]

sysuse auto, clear
reg price rep78 mpg c.rep78#c.mpg
margins, at(mpg = (10(2)40) rep78 = (1 2 3 4 5))
marginsplot

[Jakob Abekhon]

Thank you very much! This worked!