[R-br] Erro no teste de Scott-Knott

[Alisson Lucrecio]

Caro Colegas,


Boa tarde.


Estou tentando fazer um teste de Scott-Knott componente 1 (Can1) de
uma discriminante canonica, entretanto aparece o erro "There were 14
warnings (use warnings() to see them)", usando "warnings()" aparece as
mensagens abaixo.

Alguém saberia como solucionar esse problema?


Obrigado.


-- 
Alisson Lucrecio da Costa

###########------------------------------------------------------------------------###########


str(can1_scores)


'data.frame':	40 obs. of  2 variables:
 $ Tempo: Factor w/ 7 levels "0","2","4","6",..: 1 1 1 1 1 1 1 1 1 1 ...
 $ Can1 : num  -3.98 -2.08 -3.15 -2.86 -2.43 ...

can1_scores


   Tempo        Can1
1      0 -3.97903121
2      0 -2.08094673
3      0 -3.14865243
4      0 -2.86189110
5      0 -2.43198428
6      0 -2.61786107
7      0 -2.43352998
8      0 -2.40806466
9      0 -2.46111031
10     0 -3.71884011
11     2 -2.05034264
12     2 -2.00522155
13     2 -3.04292682
14     2 -1.07030352
15     2 -1.59189582
16     4 -0.48869443
17     4 -0.38234727
18     4 -0.28869622
19     4 -1.63394988
20     4 -0.38818468
21     6  1.04914435
22     6  1.35412327
23     6  0.06221095
24     6  1.72914975
25     6  1.49862908
26     8  1.88308179
27     8  3.76305713
28     8  2.61399465
29     8  3.24530432
30     8  3.28546337
31    10  2.19512933
32    10  2.71337547
33    10  2.47481491
34    10  1.49630701
35    10  1.79375735
36    12  2.24198697
37    12  2.26729170
38    12  1.57615671
39    12  2.00861523
40    12  1.83288137


av1 <- aov(Can1~Tempo, data=can1_scores)


summary(av1)


            Df Sum Sq Mean SqLevels      Means SK(5%)
      8  2.9581803      a
     10  2.1346768      b
     12  1.9853864      b
      6  1.1386515      c
      4 -0.6363745      d
      2 -1.9521381      e
      0 -2.8141912      f

There were 14 warnings (use warnings() to see them) F value Pr(>F)
Tempo        6 193.01   32.17   88.92 <2e-16 ***
Residuals   33  11.94    0.36
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1


sk1 <- summary(SK(av1, which ='Tempo'))


 Levels      Means SK(5%)
      8  2.9581803      a
     10  2.1346768      b
     12  1.9853864      b
      6  1.1386515      c
      4 -0.6363745      d
      2 -1.9521381      e
      0 -2.8141912      f
There were 14 warnings (use warnings() to see them)

warnings()


Warning messages:
1: In if ((lam < valchisq) | (ord1 == k)) { ... :
  the condition has length > 1 and only the first element will be used
2: In if ((lam < valchisq) | (ord1 == k)) { ... :
  the condition has length > 1 and only the first element will be used
3: In if ((lam < valchisq) | (ord1 == k)) { ... :
  the condition has length > 1 and only the first element will be used
4: In if (lam > valchisq) { ... :
  the condition has length > 1 and only the first element will be used
5: In if (lam < valchisq) { ... :
  the condition has length > 1 and only the first element will be used
6: In if ((lam < valchisq) | (ord1 == k)) { ... :
  the condition has length > 1 and only the first element will be used
7: In if (lam > valchisq) { ... :
  the condition has length > 1 and only the first element will be used
8: In if (lam < valchisq) { ... :
  the condition has length > 1 and only the first element will be used
9: In if ((lam < valchisq) | (ord1 == k)) { ... :
  the condition has length > 1 and only the first element will be used
10: In if (lam > valchisq) { ... :
  the condition has length > 1 and only the first element will be used
11: In if (lam < valchisq) { ... :
  the condition has length > 1 and only the first element will be used
12: In if ((lam < valchisq) | (ord1 == k)) { ... :
  the condition has length > 1 and only the first element will be used
13: In if (lam > valchisq) { ... :
  the condition has length > 1 and only the first element will be used
14: In if (lam < valchisq) { ... :
  the condition has length > 1 and only the first element will be used

###########------------------------------------------------------------------------###########
-------------- Próxima Parte ----------
Um anexo em HTML foi limpo...
URL: <http://listas.inf.ufpr.br/pipermail/r-br/attachments/20140513/b4a19a43/attachment.html>