CODEWITHSUNDEEP
rtidyr
saurabh
saurabh

Reputation: 391

join the factors in the same obervation

I have a data frame 

mat<-matrix(rnorm(100),20,5)
d<-data.frame(factor=as.factor(rep(c(1,2),each=10)),mat)

   factor          X1          X2         X3
1       1 -0.32946204  0.04911047 -1.4746314
2       1 -0.36187765 -0.36763007  0.2300900
3       1  0.77672015  1.38866125 -0.1773993
4       1  0.77633728 -1.14792986  1.2762236
5       1  0.41777137 -2.19985513 -0.5128805
6       1 -0.71586631  0.33818931  1.4241757
7       1  1.09668610 -1.33478667  0.7134252
8       1  0.03514851 -0.42269620  0.2069336
9       1  0.80135492 -0.53586509 -1.6818566
10      1 -0.60833750  1.27533884 -0.1066936
11      2  1.61718020 -0.34770915  0.6308279
12      2  4.15932386 -2.61849747  0.9735757
13      2  0.19107478 -2.32520215 -0.3536511
14      2 -0.52132006 -0.63421868  0.6935504
15      2  1.10620346 -1.40234676  1.2600821
16      2 -1.49278432 -0.09365126  0.8021205
17      2  0.31906474 -0.27821518  1.4890480
18      2 -0.94842397 -0.48807032 -0.8057988
19      2 -0.54172629  0.04712952  1.2591512
20      2 -0.43303452  0.06493442 -0.3316403

how do I convert the factors into just two observations as 1 and 2 and all the values added correspondingly

factor X1 X2 X3 X1 X2 X3 X1 X2 X3 ..........................
1          
2

i want all the observations corresponding to 1 and 2 to come in one row. I don't want sum or mean. I just want all the observations in two columns for above exampl

Upvotes: 1

Views: 39

Answers (2)

austensen
austensen

Reputation: 3017

You can do this with dplyr and tidyr.

library(dplyr)
library(tidyr)

mat <- matrix(rnorm(100),20,5)
d <- data.frame(factor=as.factor(rep(c(1,2),each=10)),mat) %>% as_data_frame()

d %>% 
  group_by(factor) %>% 
  mutate(n = row_number()) %>% 
  gather("key", "val", -factor, -n) %>% 
  unite(key_n, key, n) %>% 
  spread(key_n, val)

#> Source: local data frame [2 x 51]
#> Groups: factor [2]
#> 
#> # A tibble: 2 x 51
#>   factor      X1_1      X1_10       X1_2      X1_3       X1_4     X1_5
#> * <fctr>     <dbl>      <dbl>      <dbl>     <dbl>      <dbl>    <dbl>
#> 1      1 0.5595506 -0.9253696 -1.0211724  1.413117 -0.8939388 1.071695
#> 2      2 0.3953686 -0.6118039 -0.3091567 -1.076853 -1.4190182 2.607704
#> # ... with 44 more variables: X1_6 <dbl>, X1_7 <dbl>, X1_8 <dbl>,
#> #   X1_9 <dbl>, X2_1 <dbl>, X2_10 <dbl>, X2_2 <dbl>, X2_3 <dbl>,
#> #   X2_4 <dbl>, X2_5 <dbl>, X2_6 <dbl>, X2_7 <dbl>, X2_8 <dbl>,
#> #   X2_9 <dbl>, X3_1 <dbl>, X3_10 <dbl>, X3_2 <dbl>, X3_3 <dbl>,
#> #   X3_4 <dbl>, X3_5 <dbl>, X3_6 <dbl>, X3_7 <dbl>, X3_8 <dbl>,
#> #   X3_9 <dbl>, X4_1 <dbl>, X4_10 <dbl>, X4_2 <dbl>, X4_3 <dbl>,
#> #   X4_4 <dbl>, X4_5 <dbl>, X4_6 <dbl>, X4_7 <dbl>, X4_8 <dbl>,
#> #   X4_9 <dbl>, X5_1 <dbl>, X5_10 <dbl>, X5_2 <dbl>, X5_3 <dbl>,
#> #   X5_4 <dbl>, X5_5 <dbl>, X5_6 <dbl>, X5_7 <dbl>, X5_8 <dbl>, X5_9 <dbl>

Upvotes: 3

Roman Luštrik
Roman Luštrik

Reputation: 70653

You can split by factor, reflow the data into wide format and combine the result.

d <- split(d, f = d$factor)

out <- sapply(d, FUN = function(x) {
  x$seqi <- 1:nrow(x)
  reshape(x, idvar = "factor", timevar = "seqi", direction = "wide")
}, simplify = FALSE)

do.call(rbind, out)

  factor      X1.1       X2.1       X3.1        X4.1       X5.1
1      1 0.4954632  2.1297804 -1.0001047 -0.04452902 0.07515041
2      2 0.6782848 -0.9476608 -0.3492558  0.41029783 0.78760862
       X1.2     X2.2      X3.2      X4.2       X5.2      X1.3
1 -1.276254 -1.33387  2.153656 0.6856235 -0.6053279 0.4786269
2  1.185984  1.25427 -1.120756 0.8260916 -0.6193034 0.5087422
...

Upvotes: 2

Related Questions