Data visualization with SAS programming

Two fitted line with 3 variables

Accepted Solution Solved
Reply
Occasional Contributor syk
Occasional Contributor
Posts: 15
Accepted Solution

Two fitted line with 3 variables

How can I recreate this gplot procedure using sgplot?

symbol1 v=M i=sm70 c=black l=1;
symbol2 v=S i=sm70 c=black l=3;
proc gplot data=insurance;
   plot months*size=stock;
run;
proc sgplot data=insurance; 
	pbspline x=stock y=months*size; 

I tried using the above sgplot procedure but gave me errors because SAS UE does not recognize the "months*size" part from the former gplot procedures. I need two fitted lines within the graph because the stocks are binary 0 or 1.

Capture.GIFCapture.GIF


Accepted Solutions
Solution
a month ago
PROC Star
Posts: 551

Re: Two fitted line with 3 variables

[ Edited ]

You could also use the PBSPLINE statement as you initially intended, still group by stock and use a high value for the SMOOTH option like this

 

data insurance;
input months size stock;
datalines;
30 31  0
26 92  0
22 104 0
19 120 0
17 151 0
21 175 0
12 210 0
16 238 0
0  277 0
4  290 0
38 68  1
31 85  1
30 124 1
28 164 1
20 166 1
21 224 1
14 246 1
15 272 1
11 295 1
13 305 1
;

title 'Insurance Innovation';
title2 'With Smoothed Lines';
proc sgplot data = insurance;
   pbspline x = size y = months / group=stock smooth = 1000000;
   yaxis min=0 max=40;
   xaxis min=0 max=400;
run;

View solution in original post


All Replies
Trusted Advisor
Posts: 1,128

Re: Two fitted line with 3 variables

 

 

proc sgplot data=insurance;
  reg x=size y=month / group=stock;
run;
Thanks,
Jag
Occasional Contributor syk
Occasional Contributor
Posts: 15

Re: Two fitted line with 3 variables

What about the individual line attributes? How can I customize the lines with the group procedure? Also I need a fitted line to determine the initial correlation not a straight fitted line. The loess feature worked better.
PROC Star
Posts: 551

Re: Two fitted line with 3 variables

The REG statement supports the LINEATTRS option. See the documentation here

 

http://support.sas.com/documentation/cdl/en/grstatproc/69716/HTML/default/viewer.htm#p0mn6vl6clqbgyn...

Solution
a month ago
PROC Star
Posts: 551

Re: Two fitted line with 3 variables

[ Edited ]

You could also use the PBSPLINE statement as you initially intended, still group by stock and use a high value for the SMOOTH option like this

 

data insurance;
input months size stock;
datalines;
30 31  0
26 92  0
22 104 0
19 120 0
17 151 0
21 175 0
12 210 0
16 238 0
0  277 0
4  290 0
38 68  1
31 85  1
30 124 1
28 164 1
20 166 1
21 224 1
14 246 1
15 272 1
11 295 1
13 305 1
;

title 'Insurance Innovation';
title2 'With Smoothed Lines';
proc sgplot data = insurance;
   pbspline x = size y = months / group=stock smooth = 1000000;
   yaxis min=0 max=40;
   xaxis min=0 max=400;
run;
Occasional Contributor syk
Occasional Contributor
Posts: 15

Re: Two fitted line with 3 variables

How do I customized individual line attributes? ex) one line to be solid and one to be dotted.
☑ This topic is SOLVED.

Need further help from the community? Please ask a new question.

Discussion stats
  • 5 replies
  • 150 views
  • 0 likes
  • 3 in conversation