t function for transpose dataframe is giving wrong values-CodePudding

I have a .tsv file that I´m loaded to R with the correct parameters. This is a data frame that I want to transpose (here is the explanation: https://www.dataanalytics.org.uk/rotating-or-transposing-r-objects/). Data frame:

GCA_910579985.1_MGBC115109_genomic_prkk
OG0000000                                      20
OG0000001                                       9
OG0000002                                      13
OG0000003                                       6
OG0000004                                      14
OG0000005                                       4
GCA_910578415.1_MGBC110107_genomic_prkk
OG0000000                                      15
OG0000001                                       6
OG0000002                                       8
OG0000003                                       5
OG0000004                                       3
OG0000005                                       3

I´m using the t-function for this purpose:

tdata<-t(data)

I already force the variable data into the data frame and into a matrix just to test Then, there is an error, because all the matrix becomes 0, 1 and 2

     OG0000000       OG0000001       OG0000002       OG0000003       OG0000004       OG0000005
GCA_910579985.1_MGBC115109_genomic_prkk 0       0       0       1       1       0
GCA_910578415.1_MGBC110107_genomic_prkk 1       1       1       0       0       2

This is wrong because the original values are changing, for example, this is my expected result:

     OG0000000       OG0000001       OG0000002       OG0000003       OG0000004       OG0000005
GCA_910579985.1_MGBC115109_genomic_prkk 20       9       13       6       14       4
GCA_910578415.1_MGBC110107_genomic_prkk 15       6       8       5       3       3

So, clearly, here the function is changing the original values. Any idea or comment?

CodePudding user response：

First off, please always make your example data easily available to others by e.g. pasting the output of dput(your_data). Your data, as dataframe df are these:

df <- 
## output of dput
structure(list(ID = c("OG0000000", "OG0000001", "OG0000002", 
"OG0000003", "OG0000004", "OG0000005"), GCA_910579985.1_MGBC115109_genomic_prkk = c(15L, 6L, 8L, 5L, 3L, 3L), GCA_910578415.1_MGBC110107_genomic_prkk = c(20L, 
9L, 13L, 6L, 14L, 4L)), class = "data.frame", row.names = c(NA, 
6L))

The type conversion from numeric to character occurs upon transposition t of a dataframe. You can circumvent it as follows (df being your dataframe):

library(dplyr) ## provides the pipe operator %>%

df %>% as.matrix %>% t %>% as.data.frame

While the pipe operator %>% is not necessary, I used it to emphasise the order of transformation. It's equivalent to as.data.frame(t(as.matrix(df))).