Problem running Lasso Penelaized regression. Any suggestions

Accepted Solution Solved
Reply
Super Contributor
Posts: 303
Accepted Solution

Problem running Lasso Penelaized regression. Any suggestions

I am having issue running this  LASSO penalized regression. Could I run LASSO on categorical outcome variable?

2274  proc glmselect data=results.Aim_212 plots=all;

2275      model CSF_INT

2276  =Adiponectin

2277  Alpha_1_Antitrypsin

2278  Alpha_2_Macroglobulin

2279  Apolipoprotein_A1

2280  Apolipoprotein_CIII

2281  Apolipoprotein_H

2282  Beta_2_Microglobulin

2283  Brain_Derived_Neurotrophic_Facto

2284  CD40

2285  CD40_Ligand

2286  CSF

2287  C_Reactive_Protein

2288  Complement_3

2289  EGF

2290  Endothelin_11

2291  Eotaxin1

2292  FGF_basic1

2293  Factor_VII

2294  Ferritin

2295  Fibrinogen

2296  G_CSF

2297  Growth_Hormone

2298  Haptoglobin

2299  IFN_gamma

2300  IGF_11

2301  IL_8

2302  IL_16

2303  IL_17

2304  IL_18

2305  IL_55

2306  IL_66

2307  IL_77

2308  IL_88

2309  IL_1ra

2310  IgA

2311  IgE1

2312  IgM

2313  Insulin

2314  Leptin1

2315  Lipoprotein__a_

2316  Lymphotactin1

2317  MCP_1

2318  MDC

2319  MIP_1alpha

2320  MIP_1beta

2321  MMP_2

2322  Myeloperoxidase

2323  Myoglobin

2324  PAI_1

2325  Prostate_Specific_Antigen__Free1

2326  Prostatic_Acid_Phosphatase

2327  RANTES

2328  SAMPLE

2329  SGOT

2330  SHBG

2331  Stem_Cell_Factor

2332  TGF_alpha1

2333  TIMP_1

2334  TNF_RII

2335  Thrombopoietin

2336  Tissue_Factor1

2337  VCAM_1

2338  VEGF

2339  von_Willebrand_Factor

2340  /selection=lasso(stop=none choose=AIC);

2341  run;

ERROR: Selection aborted because there are no suitable observations for training.

NOTE: The SAS System stopped processing this step because of errors.

NOTE: There were 60 observations read from the data set RESULTS.AIM_212.

NOTE: PROCEDURE GLMSELECT used (Total process time):

      real time           0.03 seconds

      cpu time            0.03 seconds


Accepted Solutions
Solution
‎03-07-2013 08:12 AM
Respected Advisor
Posts: 2,655

Re: Problem running Lasso Penelaized regression. Any suggestions

Posted in reply to desireatem

Part of the problem here is that you have more independent variables than you have observations (64 IVs, 60 obs).  If you really want to do LASSO for continuous variables, a rule of thumb would be 4 or 5 times as many observations as you have IVs.

Hope this helps some.

Steve Denham

View solution in original post


All Replies
Solution
‎03-07-2013 08:12 AM
Respected Advisor
Posts: 2,655

Re: Problem running Lasso Penelaized regression. Any suggestions

Posted in reply to desireatem

Part of the problem here is that you have more independent variables than you have observations (64 IVs, 60 obs).  If you really want to do LASSO for continuous variables, a rule of thumb would be 4 or 5 times as many observations as you have IVs.

Hope this helps some.

Steve Denham

🔒 This topic is solved and locked.

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

Discussion stats
  • 1 reply
  • 636 views
  • 1 like
  • 2 in conversation