Spit column into many columns

Question

data = data.frame("id"=1:40,
                  "group"=1:5,
                  "score"=sample(1:4,10,r=T))
table(data[which(data$group==1),]$score)
WANT=data.frame("group"=1:5,
                  "score1"=c(0,4,0,4,0),
                  "score2"=c(4,0,0,4,0),
                  "score3"=c(0,4,0,4,0),
                  "score4"=c(0,0,4,4,0))

In data I have "score" but I want to make separate columns for each "score" and then sum up as shown here.

I also want to have my complete data frame 'WANT' and put 0s if there aren't any people but otherwise keep the same structure in terms of rows.

Possible duplicate https://stackoverflow.com/questions/8186133/faster-ways-to-calculate-frequencies-and-cast-from-long-to-wide — Ronak Shah, Jun 02 '19 at 02:46

score 0 · Answer 1 · edited Jun 02 '19 at 04:00

0

An option would be spread after getting the count or frequency by 'group' and 'score'

library(tidyverse)
data %>% 
    count(group, score) %>%
    mutate(score = str_c("score", score)) %>%
    spread(score, n, fill = 0)

If we wanted to have with the all the combinations after the count, use complete

data %>% 
  count(group, score) %>% 
  complete(group, score, fill = list(n = 0))

I also want to have my complete data frame 'WANT' and put 0s if there aren't any people but otherwise keep the same structure in terms of rows.

edited Jun 02 '19 at 04:00

bvowe

3,004
3
16
33

answered Jun 02 '19 at 02:37

akrun

874,273
37
540
662

Thanks but this just put 1s and eliminated my count values? @akrun – bvowe Jun 02 '19 at 02:42
@bvowe No, it is only `fill`ing 0's where there are no combinations, while the count values remain the same. I may be getting a different output from you because there is no set.seed while creating the data – akrun Jun 02 '19 at 02:43
Makes sense but now suppose I want to have my complete data frame 'WANT' and put 0s if there aren't any people but otherwise keep the same structure in terms of rows, does that make sense? – bvowe Jun 02 '19 at 02:51
I add to my comment that I combine this with the other question you answered good so if there is a better way to combine this with that response please kindly share. – bvowe Jun 02 '19 at 02:52
@bvowe Updated the post. But, can you please update your post with the new info as a second piece of info. – akrun Jun 02 '19 at 02:55
@akron one more question is if there is a way to efficiently combine this question with my one about the survey weight? – bvowe Jun 02 '19 at 03:29
@bvowe I updated in the other post. Also, instead of editing on the answer, can you edit on your question post – akrun Jun 02 '19 at 03:34
Error in count(., group, score) : object 'group' not found is an error i am getting now – bvowe Jun 03 '19 at 02:25
@bvowe Can you try `data %>% dplyr::count(group, score)` in case you have also loaded another package with `count` – akrun Jun 03 '19 at 03:59

Spit column into many columns

1 Answers1