I have a question about performing multiple Wilcoxon test in R. I have 7 datasets that for each I need to compare 9 different feature extraction methods using 10 classifiers. As they are stochastic methods, I have 10 values/repetitions for each classifier. A simple example for a single dataset let's assume we have 3 methods, 3classifiers, 3 repetitions:
Method, Classifier, Accuracy
M1, C1, 2.4
M1, C1, 2.3
M1, C1, 6.4
M1, C2, 0.4
M1, C2, 1.3
M1, C2, 6.8
M1, C3, 3.4
M1, C3, 2.7
M1, C3, 2.4
M2, C1, 0.4
M2, C1, 2.8
M2, C1, 3.4
M2, C2, 7.4
M2, C2, 2.3
M2, C2, 6.5
M2, C3, 1.4
M2, C3, 2.7
M2, C3, 7.4
M3, C1, 2.0
M3, C1, 1.3
M3, C1, 7.2
M3, C2, 8.4
M3, C2, 1.3
M3, C2, 9.8
M3, C3, 3.9
M3, C3, 3.7
M3, C3, 0.9
Now, I need a way to perform the Wilcoxon test between each pair of these classifiers for different methods, i.e., M1.C1 vs M2.C1,C2,C3, M1.C1 vs M3.C1,C2,C3, and so on. Can someone suggests an appropriate way to load the data and perform the multiple tests in a simple way as I am relatively new to R. Many thanks in advance for any suggestions
