Plot the Variable importance for a regression model in SAS Entreprise Guide

[chrisronaldo]

Hi

I would like to plot the selected variable in my regression model   in their importance with their coefficient. To be clear check the picture abowe to see what I mean.Can sommeone please help me  by writing the code . Thanx

[Reeza]

---

@chrisronaldo wrote:

Hi

I would like to plot the selected variable in my regression model   in their importance with their coefficient. To be clear check the picture abowe to see what I mean.Can sommeone please help me  by writing the code . Thanx

Welcome to the SAS forums. 

Writing code isn't help, it's doing your work/homework. 

If you need assistance we're happy to help however. 

Please post more details, specifically, what procedure are you using for your regression, and is that the VIF values you're plotting, the ParameterEstimates, the P-Values?  Standardized Estimate is unclear. Have you checked the plots produced by default by the PROC you're using as well? A lot of procedures have default diagnostic plots that are quite useful. 

Since this is a graphing question, I'll move this to the graphing forum, but you should also specify your SAS version, since the options for graphing have changed a lot between the versions. 

[chrisronaldo]

Hi Reeza,

Thank you for Help in advance.I just take the picture sommewhere and I would like to Plot the same for the selected varaible in my regression model.The software is SAS  EG  7.1  oder  SAS 9.4  for the programming.

I Think the VIF option in SGPLOT  have been used to generate the graphic. but I don't know how to  do it . I just guess that the standardized estimate coefficients are ploting  in their importance rank and the colors are for the signs. the blue for the positif coefficients and the red for the negative's one .Abowe are the variable use in the model.

pax_totalonboard*price_pax_weighted2015	Anzahl*price_mov_year	Anzahl*price_noi_year	Anzahl	pax_totalonboard	schall_grp	k_dow	k_month	k_seasoncode
251615,9294	228661,1746	640,556426	1276	60381	0	3	1	W
285723,9498	220776,3065	618,468273	1232	68566	0	7	1	W
277785,5529	232424,4071	651,098499	1297	66661	0	5	1	W
250103,2585	230632,3916	646,078464	1287	60018	0	3	1	W
230301,0211	210741,0199	590,356079	1176	55266	0	6	1	W
258024,9869	226510,756	634,532384	1264	61919	0	4	1	W
256895,6926	222926,7251	624,492315	1244	61648	0	1	1	W
255191,3334	227406,7638	637,042402	1269	61239	0	3	1	W
231651,1736	224539,539	629,010346	1253	55590	0	1	1	W
274160,1433	225435,5468	631,520364	1258	65791	0	7	1	W
235672,4613	225256,3452	631,01836	1257	56555	0	3	1	W
226417,2489	226689,9576	635,034388	1265	54334	0	2	1	W
277802,2214	198734,5162	556,721847	1109	66665	0	2	1	W
249861,5645	226869,1591	635,536391	1266	59960	0	4	1	W
240685,5277	223643,5313	626,500329	1248	57758	0	7	1	W
241543,9581	213429,0431	597,886131	1191	57964	0	6	1	W
266550,9502	208590,6013	584,332038	1164	63965	0	6	1	W
250174,0998	224001,9344	627,504336	1250	60035	0	2	1	W
244019,2377	224001,9344	627,504336	1250	58558	0	2	1	W
267413,5477	203035,3533	568,76993	1133	64172	0	3	1	W
213553,2955	193358,4698	541,661743	1079	51247	0	4	1	W
268709,5274	223464,3297	625,998325	1247	64483	0	7	1	W
271914,0562	228481,9731	640,054423	1275	65252	0	5	1	W
279310,7252	232603,6087	651,600502	1298	67027	0	5	1	W
248382,2307	203393,7564	569,773937	1135	59605	0	4	1	W
259858,5274	224718,7406	629,51235	1254	62359	0	1	1	W
228959,2028	221313,9112	619,974284	1235	54944	0	2	1	W
196584,7116	156801,3541	439,253035	875	47175	0	1	1	W
264154,8461	212712,2369	595,878117	1187	63390	0	5	1	W

[Reeza]

I don't think SGPLOT has a VIF option - that's for regression procedures. I didn't see answers to the relevant questions:

1. What procedure are you using for your regression?

2. Is that the VIF values you're plotting, the ParameterEstimates, or the P-Values?

[Reeza]

http://documentation.sas.com/?docsetId=statug&docsetTarget=statug_reg_examples05.htm&docsetVersion=1...

Does this example help at all?