This is a simple task, but some how I got confused. Requirement is that sort columns independent of each other. Most of us, who used to excel know that independent sorting of columns is not possible without coding. One can sort manually, but it is cumbersome when you have tons of columns. Let us look at the data what we have, then we can get onto the code.
====================================
$ df1=read.csv("sort_independent.txt", sep="\t", stringsAsFactors = F, strip.white = T)
$ df1 %>% gather(time,v, -sample) %>% arrange(time,desc(v)) %>% select (time, sample, v)
=====================================
output:
======================================
time sample v
1 Time1 sample_2 1080
2 Time1 sample_3 1032
3 Time1 sample_1 883
4 Time1 sample_5 751
5 Time1 sample_8 731
6 Time1 sample_6 493
7 Time1 sample_9 360
8 Time1 sample_10 334
9 Time1 sample_7 214
10 Time1 sample_4 60
11 Time10 sample_4 1190
12 Time10 sample_2 1127
13 Time10 sample_5 1092
14 Time10 sample_7 1008
15 Time10 sample_8 612
16 Time10 sample_9 583
17 Time10 sample_3 260
18 Time10 sample_10 78
19 Time10 sample_1 20
20 Time10 sample_6 16
===================================
This is pretty simple. However, let us this do this in a complex way so that we will have a list of data frames, one data frame for Time point and sort samples within each data frame (time point)
code:
===================================
$ df1 %>% gather(.,time,v,-sample) %>% split(.$time) %>% map(~select(.,c(1,3)) %>% arrange(desc(v)))
===================================
output:
====================================
$Time1
sample v
1 sample_2 1080
2 sample_3 1032
3 sample_1 883
4 sample_5 751
5 sample_8 731
6 sample_6 493
7 sample_9 360
8 sample_10 334
9 sample_7 214
10 sample_4 60
$Time10
sample v
1 sample_4 1190
2 sample_2 1127
3 sample_5 1092
4 sample_7 1008
5 sample_8 612
6 sample_9 583
7 sample_3 260
8 sample_10 78
9 sample_1 20
10 sample_6 16
=====================================