Rearranging data from tibble into a dataframe in R-CodePudding

I have a tibble in R arranged like this (just the first 20 rows of thousands):

A tibble: 20 × 4

	TF	confidence	target	mor
1	Ahr	B	Cyp1a1	1
2	Ahr	B	Cyp1a2	1
3	Ahr	B	Cyp1b1	1
4	Ahr	B	Fos	1
5	Ahr	B	Myc	1
6	Ahr	B	Ugt1a7c	1
7	Ahr	B	Ugt1a6a	1
8	Ahr	B	Ugt1a6b	1
9	Ahr	B	Asap1	1
10	Ahr	B	Erg	1
11	Ahr	B	Vgll4	1
12	Ar	A	Abce1	1
13	Ar	A	Acad12	1
14	Ar	A	Acad10	1
15	Ar	A	Ahsg	1
16	Ar	A	Akr1b3	1
17	Ar	A	Akt1	1
18	Ar	A	Anapc10	1
19	Ar	A	Ap2m1	1
20	Ar	A	Appbp2	1

and I want to arrange it to look like this:

TF	Description	Targets
Ahr	None	Cyp1a1	Cyp1a2	Cyp1b1	Fos	Myc	Ugt1a7c	Ugt1a6a	Ugt1a6b	Asap1	Erg	Vgll4
Ar	None	Abce1	Acad12	Acad10	Ahsg	Akr1b3	Akr1b3	Akt1	Anapc10	Ap2m1	Appbp2

...The table goes on for thousands of lines. There can be a varying number of targets for a given TF value. How do I create this second table as a dataframe in R given the first table?

Thank you!

CodePudding user response：

library(dplyr)

df %>% 
  group_by(TF) %>% 
  summarise(Targets = list(target))

#>    TF                                                                        Targets
#> 1 Ahr Cyp1a1, Cyp1a2, Cyp1b1, Fos, Myc, Ugt1a7c, Ugt1a6a, Ugt1a6b, Asap1, Erg, Vgll4
#> 2  Ar              Abce1, Acad12, Acad10, Ahsg, Akr1b3, Akt1, Anapc10, Ap2m1, Appbp2

This results in a so-called list-column (Targets is a list of character vectors), that will work for varying numbers of targets per TF.

With data:

df <- read.table(text =
'TF confidence  target  mor
Ahr B   Cyp1a1  1
Ahr B   Cyp1a2  1
Ahr B   Cyp1b1  1
Ahr B   Fos 1
Ahr B   Myc 1
Ahr B   Ugt1a7c 1
Ahr B   Ugt1a6a 1
Ahr B   Ugt1a6b 1
Ahr B   Asap1   1
Ahr B   Erg 1
Ahr B   Vgll4   1
Ar  A   Abce1   1
Ar  A   Acad12  1
Ar  A   Acad10  1
Ar  A   Ahsg    1
Ar  A   Akr1b3  1
Ar  A   Akt1    1
Ar  A   Anapc10 1
Ar  A   Ap2m1   1
Ar  A   Appbp2  1',
header = TRUE)