i mann-whitney u test comparing columns 2,3,4 against columns 5, 6 , 7 , p-value each row assess significance.
type sp1 sp2 sp3 sp4 sp5 sp6 gh1 4 4 4 2 2 0 gh2 7 7 7 4 8 4 gh3 13 17 16 10 16 10 gh5 9 10 10 11 10 6 gh6 0 0 0 0 0 0 gh7 1 1 1 1 1 0 gh9 0 0 1 0 1 0 gh10 3 1 1 0 2 0 gh11 1 1 0 2 1 1 gh12 3 1 2 3 1 2 gh13 11 13 12 12 13 12 gh15 2 1 2 2 1 3 gh16 17 16 16 14 11 13 gh17 7 7 7 8 6 7 gh18 7 9 9 9 9 5 gh20 2 2 2 2 1 2 gh22 0 0 0 0 0 0 gh23 2 1 2 1 0 3 gh24 0 0 0 1 0 0 gh25 1 1 1 1 0 0 gh26 0 0 0 0 0 0 i did see tutorials comparing 2 sets of values
wilcox.test(a,b, correct=false) but not sure how 2 groups , p value each row know if of types significant across 2 groups. in end
type sp1 sp2 sp3 sp4 sp5 sp6 p-value gh1 4 4 4 2 2 0 0.4 gh2 7 7 7 4 8 4 0.2 gh3 13 17 16 10 16 10 0.01 gh5 9 10 10 11 10 6 0.6
df$p_values <- apply(df, 1, function(x) { wilcox.test(as.numeric(x[2:4]), as.numeric(x[5:7]))$p.value }) type sp1 sp2 sp3 sp4 sp5 sp6 p_values 1 gh1 4 4 4 2 2 0 0.05934644 2 gh2 7 7 7 4 8 4 0.63735189 3 gh3 13 17 16 10 16 10 0.26115456 4 gh5 9 10 10 11 10 6 1.00000000 5 gh6 0 0 0 0 0 0 nan 6 gh7 1 1 1 1 1 0 0.50498508 7 gh9 0 0 1 0 1 0 1.00000000 8 gh10 3 1 1 0 2 0 0.36868827 9 gh11 1 1 0 2 1 1 0.30169958 10 gh12 3 1 2 3 1 2 1.00000000 11 gh13 11 13 12 12 13 12 0.81366372 12 gh15 2 1 2 2 1 3 0.81366372 13 gh16 17 16 16 14 11 13 0.07652250 14 gh17 7 7 7 8 6 7 1.00000000 15 gh18 7 9 9 9 9 5 1.00000000 16 gh20 2 2 2 2 1 2 0.50498508 17 gh22 0 0 0 0 0 0 nan 18 gh23 2 1 2 1 0 3 0.82218677 19 gh24 0 0 0 1 0 0 0.50498508 20 gh25 1 1 1 1 0 0 0.18763233 21 gh26 0 0 0 0 0 0 nan 
Comments
Post a Comment