I need to run non-parametric test and posthoc test on my data, I was using PMCMRPlus package, but after upgrading MacOS to Ventura RStudio doesn't run anymore. Does anyone have any suggestions for another package for this posthoc test?
Error: 'posthoc.kruskal.nemenyi.test.default' is defunct.
Use 'PMCMRplus::kwAllPairsNemenyiTest' instead.
See help("Defunct") and help("PMCMR-defunct").
library(PMCMRplus)
ans <- posthoc.kruskal.nemenyi.test.default(count ~ spray, data = InsectSprays)
#> Error in posthoc.kruskal.nemenyi.test.default(count ~ spray, data = InsectSprays): could not find function "posthoc.kruskal.nemenyi.test.default"
ans <- kwAllPairsNemenyiTest(count ~ spray, data = InsectSprays)
#> Warning in kwAllPairsNemenyiTest.default(c(10, 7, 20, 14, 14, 12, 10, 23, : Ties
#> are present, p-values are not corrected.
summary(ans)
#>
#> Pairwise comparisons using Tukey-Kramer-Nemenyi all-pairs test with Tukey-Dist approximation
#> data: count by spray
#> alternative hypothesis: two.sided
#> P value adjustment method: single-step
#> H0
#> q value Pr(>|q|)
#> B - A == 0 0.441 0.99960707
#> C - A == 0 6.738 2.7914e-05 ***
#> D - A == 0 4.400 0.02292762 *
#> E - A == 0 5.435 0.00169497 **
#> F - A == 0 0.572 0.99861029
#> C - B == 0 7.179 5.7061e-06 ***
#> D - B == 0 4.841 0.00812908 **
#> E - B == 0 5.876 0.00046606 ***
#> F - B == 0 0.131 0.99999905
#> D - C == 0 2.338 0.56300179
#> E - C == 0 1.303 0.94108652
#> F - C == 0 7.311 3.4955e-06 ***
#> E - D == 0 1.035 0.97809294
#> F - D == 0 4.973 0.00584731 **
#> F - E == 0 6.007 0.00031142 ***
#> ---
#> Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Created on 2023-01-21 with reprex v2.0.2
You should
update.packages("PMCMRplus")
update R to v.4.1 or 4.2 and run from the R.App.
Thank you so much!
but now I have another problem, kruskal.test is significant and the posthoc test doesn't shows difference. Could it be due to the message that "p-values are not corrected" using kwAllPairsNemenyiTest?
kruskal.test(DESEMP$TCE~trat)
Kruskal-Wallis rank sum test
data: DESEMP$TCE by trat
Kruskal-Wallis chi-squared = 26.446, df = 14, p-value = 0.02271
ansTCE <- kwAllPairsNemenyiTest(DESEMP$TCE~trat, dist = c("Tukey", "Chisquare"))
Warning message:
In kwAllPairsNemenyiTest.default(c(71.07142857, 32.88095238, -33.19047619, :
Ties are present, p-values are not corrected.
summary(ansTCE)
Pairwise comparisons using Tukey-Kramer-Nemenyi all-pairs test with Tukey-Dist approximation
Signif. codes: 0 ‘’ 0.001 ‘’ 0.01 ‘’ 0.05 ‘.’ 0.1 ‘ ’ 1
signifies that there were tied ranks. For an explanation of how this affects the way in which test statistics are calculated, see this paper by the package author.
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.
If you have a query related to it or one of the replies, start a new topic and refer back with a link.