CODEWITHSUNDEEP
rp-value
Lara
Lara

Reputation: 119

Why do I have different p-values if calculated differently?

I do have the following dataframe:

df <- data.frame(
  Assay = c("Gene1", "Gene2", "Gene3"),
  DT1 = c(1,2,3),
  DT2 = c(4,5,6),
  DT3 = c(4,5,6),
  DT4 = c(0,8,7),
  DT5 = c(-1,2,5),
  DT6 = c(4,5,3),
  DT7 = c(5,2,9),
  DT8 = c(0,0,4),
  DT9 = c(3,6,2),
  DT10 = c(5,9,1),
  DT11 = c(2,3,4),
  DT12 = c(8,1,6)
)

And I would like to create a column that will contain p-values for groups compared row by row. First 5 columns (2:6) versus Next 7 columns (7:13)

# Perform t-tests row-wise and obtain p-values
p_values <- apply(df[, 2:6], 1, function(row) {
  t_test_result <- t.test(row, df[, 7:13])
  t_test_result$p.value
})

# Add the p-values column to the dataframe
df$p_values <- p_values
df

For the first row when I use this script I have a p-value of 0.09335115 while if I do it manually:

t.test(c(1,4,4,0,-1),
       c(4,5,0,3,5,2,8))

I do have a p-value of 0.1425 What's the issue?

Upvotes: 0

Views: 64

Answers (1)

Onyambu
Onyambu

Reputation: 79238

just do:

p_values <- apply(df[-1], 1, function(row) {
     t_test_result <- t.test(row[1:5], row[6:12])
     t_test_result$p.value
 })
 
p_values
[1] 0.1425172 0.6840726 0.3266262

Upvotes: 2

Related Questions