Rows to Columns

library(tidyverse)

## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.1 ──

## ✓ ggplot2 3.3.5     ✓ purrr   0.3.4
## ✓ tibble  3.1.4     ✓ dplyr   1.0.7
## ✓ tidyr   1.1.3     ✓ stringr 1.4.0
## ✓ readr   2.0.1     ✓ forcats 0.5.1

## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## x dplyr::filter() masks stats::filter()
## x dplyr::lag()    masks stats::lag()

Let’s say I want data of the following shape:

example <- tribble(
  ~gene, ~time0h, ~time1h, ~time3h, ~time4h,
  "ASS", 4.12, 5.2, 3.4, 3.2,
  "BAS", 1.2, 3.4, 4.6, 3.8,
  "CAS", 10.2, 2.3, 0.8, 0.1
  
)

example

## # A tibble: 3 × 5
##   gene  time0h time1h time3h time4h
##   <chr>  <dbl>  <dbl>  <dbl>  <dbl>
## 1 ASS     4.12    5.2    3.4    3.2
## 2 BAS     1.2     3.4    4.6    3.8
## 3 CAS    10.2     2.3    0.8    0.1

(Example from Mfuzz)

To start I have the following shape:

df <- tribble(
  ~time, ~ASS, ~BAS, ~CAS,
  "time0h", 4.12, 1.2, 10.2,
  "time1h", 5.2, 3.4, 2.3,
  "time3h", 3.4, 4.6, 0.8,
  "time4h", 3.2, 3.8, 0.1
)

df

## # A tibble: 4 × 4
##   time     ASS   BAS   CAS
##   <chr>  <dbl> <dbl> <dbl>
## 1 time0h  4.12   1.2  10.2
## 2 time1h  5.2    3.4   2.3
## 3 time3h  3.4    4.6   0.8
## 4 time4h  3.2    3.8   0.1

So I want to transpose the dataframe, since my rows are what I want my columns to be:

t(df)

##      [,1]     [,2]     [,3]     [,4]    
## time "time0h" "time1h" "time3h" "time4h"
## ASS  "4.12"   "5.20"   "3.40"   "3.20"  
## BAS  "1.2"    "3.4"    "4.6"    "3.8"   
## CAS  "10.2"   " 2.3"   " 0.8"   " 0.1"

While the code is simple, it creates a character vector:

typeof(t(df))

## [1] "character"

And I have to do some manual clean up when converting back to a tibble as a result:

reshaped <- df %>%
  select(!time) %>%
  t() %>%
  as_tibble(.name_repair = "minimal")

colnames(reshaped) <- df$time
reshaped$gene <- colnames(df)[2:ncol(df)]

reshaped <- reshaped %>%
  relocate(gene)

reshaped

## # A tibble: 3 × 5
##   gene  time0h time1h time3h time4h
##   <chr>  <dbl>  <dbl>  <dbl>  <dbl>
## 1 ASS     4.12    5.2    3.4    3.2
## 2 BAS     1.2     3.4    4.6    3.8
## 3 CAS    10.2     2.3    0.8    0.1

It feels like I should be able to do a transpose without getting the wonky intermediate character vector, but couldn’t find a simple solution with either pivot_* from {tidyr}.