CODEWITHSUNDEEP
rtidyrdplyr
azizi tamimi
azizi tamimi

Reputation: 75

Tidying data using R

I have a problem tidying a large dataset that would highly appreciate your help with. A subset of my data would look like this: 

> d1
# A tibble: 16 x 8
   Subject   RT1   RT2 item                 ROI `Item _Number` IA_LABEL  block
     <dbl> <dbl> <dbl> <chr>              <dbl>          <dbl> <chr>     <dbl>
 1       1   567   261 brilliant salesman     1              1 brilliant     3
 2       1   494    76 brilliant salesman     2              1 salesman      3
 3       1   441   211 detailed brochure      1              2 detailed      3
 4       1   544   282 detailed brochure      2              2 brochure      3
 5       1   289   153 detailed brochure      1              2 detailed      4
 6       1   141   141 detailed brochure      2              2 brochure      4
 7       1   141   141 brilliant salesman     1              1 brilliant     4
 8       1   263    90 brilliant salesman     2              1 salesman      4
 9       2   216   216 detailed brochure      1              2 detailed      3
10       2   248   248 detailed brochure      2              2 brochure      3
11       2   216   128 brilliant salesman     1              1 brilliant     3
12       2   238   140 brilliant salesman     2              1 salesman      3
13       2   212   212 detailed brochure      1              2 detailed      6
14       2   369   241 detailed brochure      2              2 brochure      6
15       2   208   208 brilliant salesman     1              1 brilliant     6
16       2   191   191 brilliant salesman     2              1 salesman      6

dput (d1)
structure(list(Subject = c(1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 
2, 2, 2, 2), RT1 = c(567, 494, 441, 544, 289, 141, 141, 263, 
216, 248, 216, 238, 212, 369, 208, 191), RT2 = c(261, 76, 211, 
282, 153, 141, 141, 90, 216, 248, 128, 140, 212, 241, 208, 191
), item = c("brilliant salesman", "brilliant salesman", "detailed brochure", 
"detailed brochure", "detailed brochure", "detailed brochure", 
"brilliant salesman", "brilliant salesman", "detailed brochure", 
"detailed brochure", "brilliant salesman", "brilliant salesman", 
"detailed brochure", "detailed brochure", "brilliant salesman", 
"brilliant salesman"), ROI = c(1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 
1, 2, 1, 2, 1, 2), `Item _Number` = c(1, 1, 2, 2, 2, 2, 1, 1, 
2, 2, 1, 1, 2, 2, 1, 1), IA_LABEL = c("brilliant", "salesman", 
"detailed", "brochure", "detailed", "brochure", "brilliant", 
"salesman", "detailed", "brochure", "brilliant", "salesman", 
"detailed", "brochure", "brilliant", "salesman"), block = c(3, 
3, 3, 3, 4, 4, 4, 4, 3, 3, 3, 3, 6, 6, 6, 6)), row.names = c(NA, 
-16L), class = c("tbl_df", "tbl", "data.frame"))

This data represents the reaction times (RT1 and RT2) once reading some words (IA_Label). In each row of the column (IA_LABEL), is the reaction times (RT1 AND RT2) for only ONE word. So, these words in IA_LABEL represents the first word of a phrase (where ROI=1) and the second word of the same phrase (where ROI=2). The way I programmed this experiment allowed me to only look at each word separately. I am trying also to look at RT1 and RT2 for the whole phrase as well (RT1 for the whole phrase is RT1 where ROI is 1 + RT1 where ROI is 2. The same process can be applied for the RT2 by summing its values where ROI =1 + where ROI=2.
I am trying to create two columns (variables) by summing the values of the dependent variables (RT1 and RT2).

So, ideally I am hoping to have a code that would mutate two variables in a similar way to the one below, where the first row of Sum_RT1 = RT1 of that row (where ROI=1) + RT2 of the subsequent row where ROI=2. The second row of Sum_RT1 would be the same value since the same summing process would be applied, but in the reverse direction (i.e., RT1 where ROI is 2 + RT1 where ROI is 2. 

> d2
# A tibble: 16 x 10
   Subject   RT1   RT2 item     ROI `Item _Number` IA_LABEL block Sum_RT1 Sum_RT2
     <dbl> <dbl> <dbl> <chr>  <dbl>          <dbl> <chr>    <dbl>   <dbl>   <dbl>
 1       1   567   261 brill~     1              1 brillia~     3    1061     337
 2       1   494    76 brill~     2              1 salesman     3    1061     337
 3       1   441   211 detai~     1              2 detailed     3     985     493
 4       1   544   282 detai~     2              2 brochure     3     985     493
 5       1   289   153 detai~     1              2 detailed     4     430     294
 6       1   141   141 detai~     2              2 brochure     4     430     294
 7       1   141   141 brill~     1              1 brillia~     4     404     231
 8       1   263    90 brill~     2              1 salesman     4     404     231
 9       2   216   216 detai~     1              2 detailed     3     464     464
10       2   248   248 detai~     2              2 brochure     3     464     464
11       2   216   128 brill~     1              1 brillia~     3     454     268
12       2   238   140 brill~     2              1 salesman     3     454     268
13       2   212   212 detai~     1              2 detailed     6     581     453
14       2   369   241 detai~     2              2 brochure     6     581     453
15       2   208   208 brill~     1              1 brillia~     6     399     399
16       2   191   191 brill~     2              1 salesman     6     399     399

I plan to apply the the process for many variables at a larger data, so would highly apprecaite your thoughts and help.

Upvotes: 0

Views: 83

Answers (1)

Em Laskey
Em Laskey

Reputation: 607

Would grouping the data by Subject, Item_Number & block give the desired output? 

colnames(d1)[6] <- "Item_Number" # delete the blank in the column name (otherwise it gives me an error...)

d2 <- d1 %>%
  group_by(Subject,block, Item_Number) %>%
  mutate(SumRT1 = sum(RT1), 
         SumRT2 = sum(RT2))

Output:

# A tibble: 16 x 10
# Groups:   Subject, block, Item_Number [8]
   Subject   RT1   RT2 item                 ROI Item_Number IA_LABEL  block SumRT1 SumRT2
     <dbl> <dbl> <dbl> <chr>              <dbl>       <dbl> <chr>     <dbl>  <dbl>  <dbl>
 1       1   567   261 brilliant salesman     1           1 brilliant     3   1061    337
 2       1   494    76 brilliant salesman     2           1 salesman      3   1061    337
 3       1   441   211 detailed brochure      1           2 detailed      3    985    493
 4       1   544   282 detailed brochure      2           2 brochure      3    985    493
 5       1   289   153 detailed brochure      1           2 detailed      4    430    294
 6       1   141   141 detailed brochure      2           2 brochure      4    430    294
 7       1   141   141 brilliant salesman     1           1 brilliant     4    404    231
 8       1   263    90 brilliant salesman     2           1 salesman      4    404    231
 9       2   216   216 detailed brochure      1           2 detailed      3    464    464
10       2   248   248 detailed brochure      2           2 brochure      3    464    464
11       2   216   128 brilliant salesman     1           1 brilliant     3    454    268
12       2   238   140 brilliant salesman     2           1 salesman      3    454    268
13       2   212   212 detailed brochure      1           2 detailed      6    581    453
14       2   369   241 detailed brochure      2           2 brochure      6    581    453
15       2   208   208 brilliant salesman     1           1 brilliant     6    399    399
16       2   191   191 brilliant salesman     2           1 salesman      6    399    399

Upvotes: 1

Related Questions