Create new data set based on regression intercept and slope

[jonatan_velarde]

Good morning SAS friends:

I have this data set:

data have;
input box length weight ;
cards;
1 8.1 6.3
1 9.1 9.6
1 10.2 11.6
1 11.9 18.5
1 12.2 26.2
1 13.8 36.1
1 14.8 40.1
1 15.7 47.3
2 16.6 65.6
2 17.7 69.4
2 18.7 76.4
2 19 82.5
2 20.6 106.6
2 21.9 119.8
2 22.9 169.2
2 23.5 173.3
;

using this data set, i apply two regressions, one for box 1 and other for box 2, resulting some tables, the first one for ANOVA and the second one indicating the estimates of the variables, the last of the is the main reason to note:

For box 1:

Parameter Estimates
Variable	DF	Parameter Estimate	Standard Error	t Value	Pr > |t|
Intercept	1	-42.10680	5.35822	-7.86	0.0002
length	1	5.55902	0.43770	12.70	<.0001

and for Box 2:

Parameter Estimates
Variable	DF	Parameter Estimate	Standard Error	t Value	Pr > |t|
Intercept	1	-224.06587	41.01922	-5.46	0.0016
length	1	16.50296	2.02574	8.15	0.0002

The main objective of this problem is create a new data set containing both the intercept and slope for each box, like this:

Box	intercept	length
1	-42.1068	5.55902
2	-224.06587	16.50296

Thanks in advance

[PaigeMiller · Accepted solution]

proc reg data=have outest=parameters;
by box;
model weight=length;
run;

[Ksharp]

data have; input box length weight ; cards; 1 8.1 6.3 1 9.1 9.6 1 10.2 11.6 1 11.9 18.5 1 12.2 26.2 1 13.8 36.1 1 14.8 40.1 1 15.7 47.3 2 16.6 65.6 2 17.7 69.4 2 18.7 76.4 2 19 82.5 2 20.6 106.6 2 21.9 119.8 2 22.9 169.2 2 23.5 173.3 ; proc reg data=have outest=want noprint; by box; model weight=length; quit;

[Ksharp]

data have; input box length weight ; cards; 1 8.1 6.3 1 9.1 9.6 1 10.2 11.6 1 11.9 18.5 1 12.2 26.2 1 13.8 36.1 1 14.8 40.1 1 15.7 47.3 2 16.6 65.6 2 17.7 69.4 2 18.7 76.4 2 19 82.5 2 20.6 106.6 2 21.9 119.8 2 22.9 169.2 2 23.5 173.3 ; proc reg data=have outest=want noprint; by box; model weight=length; quit;