[R] post hoc test to a significant Kruskal-Wallis test package asbio function pairw.kw()

[Andrew Koeser]

Luigi,

agricolae has a nice function for this with the post hoc test as part of 
the output.

AKK

On 10/21/2013 11:12 PM, arun wrote:
> Hi,
> Try:
> pairw.kw(y,factor(x),conf=.95)
>
>
> A.K.
>
>
>
>
> On Monday, October 21, 2013 7:10 PM, luigi marongiu <marongiu.luigi at gmail.com> wrote:
> Dear all,
> I am trying to apply a post hoc test to a significant Kruskal-Wallis test.
> I eventually found the function ‘pairw.kw()’ of the package “asbio”. This
> function has the following syntax:
> pairw.kw(y, x, conf)
> Arguments
> y     The response variable. A vector of quantitative responses.
> x     An explanatory variable. A vector of factor levels.
> conf     The level of desired confidence, 1 - P(type I error).
>
> The example reported indicated 2 vectors of length 12 arranged in a data
> frame:
> rye.data <- data.frame(rye = c(50, 49.8, 52.3, 44.5, 62.3, 74.8, 72.5, 80.2,
> 47.6, 39.5, 47.7,50.7), nutrient = factor(c(rep(1, 4), rep(2, 4), rep(3,
> 4))))
> with(rye.data, pairw.kw(y = rye, x = nutrient, conf = .95))
>
> I tried to arrange the data in this way, as well as using my own dataframe
> obtained from the read.table() function and directly with two vectors, but
> I always obtained an error. Essentially the two vectors are not of the same
> length; this is incorrect since there are both 64 elements.
>
> Is there somebody who could tell me how to use this function?
>
> Many thanks,
> Luigi
>
> EXAMPLE GIVEN
> Here I am using the vector example.
>
> ### define vectors
> x<-c(    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
> 0,    1,    1,    1,    1,    1,    1,    1,    1,    1,    1,    1,
> 1,    2,    2,    2,    2,    2,    2,    2,    2,    2,    2,    2,
> 2,    2,    2,    2,    2,    3,    3,    3,    3,    3,    3,    3,
> 3,    3,    3,    3,    3,    3,    3,    3,    4,    4,    4,    4,
> 4,    4,    4,    4,    4)
>
> y<-c(    0.002515324,    0.005128344,    0.008474943,    0.009992525,
> 0.01460982,    0.020567249,    0.069597233,    0.192428254,
> 0.636910481,    0.673830995,    4.354055891,    4.847755691,
> 0.403135211,    0.539626363,    0.892722111,    0.993636963,
> 1.289747804,    3.766986726,    5.623002684,    8.52186547,
> 17.24684431,    19.84624792,    25.9529251,    35.64632645,
> 0.437492775,    0.594863773,    0.64377689,    0.88033664,
> 3.199906591,    5.745723357,    6.339129057,    6.583851172,
> 11.27921212,    13.01230334,    14.61151433,    26.68041064,
> 52.42360789,    71.88062442,    91.34132781,    115.1038548,
> 0.059333604,    0.068746428,    0.1473799,    0.220983138,
> 0.959677864,    1.21197612,    2.860983172,    4.402992963,
> 5.122748306,    8.711530662,    25.6245157,    44.0798333,
> 45.24736747,    66.4490478,    86.60427738,    0.10348061,
> 0.390758249,    0.634667258,    1.122594755,    2.726868877,
> 4.15194739,    5.361175006,    8.825635885,    45.75640999)
>
> ### confirm length:
>      length(x)    # 64
>      length(y)    # 64
>
> ### KRUSKAL-WALLIS to test inter-group differences
>      kruskal.test(y, x)     # p-value = 0.0008701
>
> ### Multiple pairwise comparison procedure to accompany a Kruskal-Wallis
> test
>      pairw.kw(y, x, conf = .95)
>
> # Error in outer(levels(x), levels(x), function(x1, x2) { :
> # dims [product 0] do not match the length of object [1]
>
>      [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.