Regression Screen

!SCREEN and !SMX

!SCREEN [n] [ !SMX m]
performs a 'Regression Screen', a form of all subsets regression. It produces a report to the .asr file showing the sums of squares attributable to each submodel including up to six of up to 20 terms in the DENSE equations. For example, which three of my set of 12 covariates together best explain my dependent variable given the other terms in the model. Use the !DENSE qualifier to control which terms are screened. The screen is conditional on all other terms (those in the SPARSE equations) being present. Every submodel will have at least n terms [default 1].

Use !SMX m to set the maximum number of terms to include in each model [default 6]. !SCREEN is intended for use when the candidates for inclusion in the model are main effects - e.g. a set of covariates - rather than of higher order (interactions).

Example

Finally we display a portion of Regression Screen output. The qualifier was !SCREEN 3 !SMX 3
  Source                Model  terms     Gamma     Component    Comp/SE   % C
  idsize                   92     92  0.581102      0.136683       3.31   0 P
  expt.idsize             828    828  0.121231      0.285153E-01   1.12   0 P
  Variance                504    438   1.00000      0.235214      12.70   0 P

                                    Wald F statistics
      Source of Variation           NumDF    DenDF-con F-add    F-con M P-con
  113 mu                                1      72.4 65452.25     NA   .    NA
    2 expt                              6      37.5     5.27     0.64 A 0.695

    4 type                              4      63.8    22.95     3.01 A 0.024
  114 expt.type                        10      79.3     1.31     0.93 B 0.508
   23 x20                               1      55.1     4.33     2.37 B 0.130
   24 x21                               1      63.3     1.91     0.87 B 0.355
   25 x23                               1      68.3    23.93     0.11 B 0.745

   26 x39                               1      79.7     1.85     0.35 B 0.556
   27 x48                               1      69.9     1.58     2.08 B 0.154
   28 x59                               1      49.7     1.41     0.08 B 0.779
   29 x60                               1      59.6     1.46     0.42 B 0.518
   30 x61                               1      64.0     1.11     0.04 B 0.838

   31 x62                               1      61.8     2.18     0.09 B 0.770
   32 x64                               1      55.6    31.48     4.50 B 0.038
   33 x65                               1      57.8     4.72     6.12 B 0.016
   34 x66                               1      58.5     1.13     0.03 B 0.872
   35 x70                               1      59.3     1.71     1.40 B 0.242

   36 x71                               1      64.4     0.08     0.01 B 0.929
   37 x73                               1      59.0     1.79     3.01 B 0.088
   38 x75                               1      59.9     0.04     0.26 B 0.613
   39 x91                               1      63.8     1.44     1.44 B 0.234
  Notice: The DenDF values are calculated ignoring fixed/boundary/singular
              variance parameters using empirical derivatives.

  129 mvetimates                          9 effects fitted
    9 idsize                               92 effects fitted (       7 are zero)
  115 expt.idsize                         828 effects fitted (     672 are zero)
  127 at(expt,6).type.idsize.meth           9 effects fitted (+    2199 singular)
  128 at(expt,7).type.idsize.meth          10 effects fitted (+    2198 singular)

 LINE REGRESSION      RESIDUAL             ADJUSTED  FACTORS INCLUDED
  NO DF SUMSQUARES    DF MEANSQU R-SQUARED R-SQUARED 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23
   1  3  0.1113D+02  452  0.2460  0.09098   0.08495   1  1  1  0  0  0  0  0  0  0  0  0  0  0  0  0  0
                                    *****     *****
   2  3  0.1180D+02  452  0.2445  0.09648   0.09049   1  0  1  1  0  0  0  0  0  0  0  0  0  0  0  0  0
                                    *****     *****
   3  3  0.1843D+01  452  0.2666  0.01507   0.00853   0  1  1  1  0  0  0  0  0  0  0  0  0  0  0  0  0
   4  3  0.1095D+02  452  0.2464  0.08957   0.08353   1  1  0  1  0  0  0  0  0  0  0  0  0  0  0  0  0
   5  3  0.1271D+02  452  0.2425  0.10390   0.09795   1  0  0  1  1  0  0  0  0  0  0  0  0  0  0  0  0
                                    *****     *****
   6  3  0.9291D+01  452  0.2501  0.07594   0.06981   0  1  0  1  1  0  0  0  0  0  0  0  0  0  0  0  0
   7  3  0.9362D+01  452  0.2499  0.07652   0.07039   0  0  1  1  1  0  0  0  0  0  0  0  0  0  0  0  0
   8  3  0.1357D+02  452  0.2406  0.11091   0.10501   1  0  1  0  1  0  0  0  0  0  0  0  0  0  0  0  0
                                    *****     *****
   9  3  0.9404D+01  452  0.2498  0.07687   0.07074   0  1  1  0  1  0  0  0  0  0  0  0  0  0  0  0  0
  10  3  0.1266D+02  452  0.2426  0.10350   0.09755   1  1  0  0  1  0  0  0  0  0  0  0  0  0  0  0  0
  11  3  0.1261D+02  452  0.2427  0.10313   0.09717   1  0  0  0  1  1  0  0  0  0  0  0  0  0  0  0  0
  12  3  0.9672D+01  452  0.2492  0.07906   0.07295   0  1  0  0  1  1  0  0  0  0  0  0  0  0  0  0  0
  13  3  0.9579D+01  452  0.2494  0.07830   0.07218   0  0  1  0  1  1  0  0  0  0  0  0  0  0  0  0  0
  14  3  0.9540D+01  452  0.2495  0.07797   0.07185   0  0  0  1  1  1  0  0  0  0  0  0  0  0  0  0  0
  15  3  0.1089D+02  452  0.2465  0.08907   0.08302   1  0  0  1  0  1  0  0  0  0  0  0  0  0  0  0  0
  16  3  0.2917D+01  452  0.2642  0.02384   0.01736   0  1  0  1  0  1  0  0  0  0  0  0  0  0  0  0  0
  17  3  0.2248D+01  452  0.2657  0.01838   0.01187   0  0  1  1  0  1  0  0  0  0  0  0  0  0  0  0  0
  18  3  0.1111D+02  452  0.2460  0.09088   0.08484   1  0  1  0  0  1  0  0  0  0  0  0  0  0  0  0  0
  19  3  0.1746D+01  452  0.2668  0.01427   0.00773   0  1  1  0  0  1  0  0  0  0  0  0  0  0  0  0  0
  20  3  0.1030D+02  452  0.2478  0.08423   0.07815   1  1  0  0  0  1  0  0  0  0  0  0  0  0  0  0  0
  21  3  0.1279D+02  452  0.2423  0.10454   0.09860   1  0  0  0  0  1  1  0  0  0  0  0  0  0  0  0  0
  22  3  0.8086D+01  452  0.2527  0.06609   0.05989   0  1  0  0  0  1  1  0  0  0  0  0  0  0  0  0  0
  23  3  0.7437D+01  452  0.2542  0.06079   0.05456   0  0  1  0  0  1  1  0  0  0  0  0  0  0  0  0  0
  24  3  0.1071D+02  452  0.2469  0.08755   0.08149   0  0  0  1  0  1  1  0  0  0  0  0  0  0  0  0  0
  25  3  0.1370D+02  452  0.2403  0.11200   0.10611   0  0  0  0  1  1  1  0  0  0  0  0  0  0  0  0  0
                                    *****     *****
  26  3  0.1511D+02  452  0.2372  0.12351   0.11770   1  0  0  0  1  0  1  0  0  0  0  0  0  0  0  0  0
                                    *****     *****
  27  3  0.1353D+02  452  0.2407  0.11064   0.10473   0  1  0  0  1  0  1  0  0  0  0  0  0  0  0  0  0
 ...
 680  3  0.1057D+02  452  0.2472  0.08641   0.08035   1  1  0  0  0  0  0  0  0  0  0  0  0  0  0  0  1

Return to start