I am looking for a way to transfor a data.frame with thousands of rows like this
code date value uname tcode
<chr> <date> <dbl> <ord> <int>
1 CODE1 1968-02-01 14.1 "" NA
2 CODE1 1968-03-01 9.50 "" NA
3 CODE1 1968-04-01 22.1 "" NA
4 CODE2 1968-02-01 15.1 "" NA
5 CODE2 1968-03-01 13.50 "" NA
6 CODE2 1968-04-01 23.1 "" NA
7 CODE3 1968-02-01 16.1 "" NA
8 CODE3 1968-03-01 15.50 "" NA
9 CODE3 1968-04-01 13.1 "" NA
Into something like:
date CODE1 CODE2 CODE3
<date> <dbl> <dbl> <dbl>
1 1968-02-01 14.1 15.1 16.1
2 1968-03-01 9.50 13.50 15.50
3 1968-04-01 22.1 23.1 13.1
This seems straightforward but I am having difficulty realizing this task. Thanks!
With tidyverse you can use pivot_wider
library(dplyr)
library(tidyr)
df %>% select(-c(uname,tcode)) %>% pivot_wider(names_from="code")
# A tibble: 3 x 4
date CODE1 CODE2 CODE3
<chr> <dbl> <dbl> <dbl>
1 1968-02-01 14.1 15.1 16.1
2 1968-03-01 9.5 13.5 15.5
3 1968-04-01 22.1 23.1 13.1
df <- structure(list(code = c("CODE1", "CODE1", "CODE1", "CODE2", "CODE2",
"CODE2", "CODE3", "CODE3", "CODE3"), date = c("1968-02-01", "1968-03-01",
"1968-04-01", "1968-02-01", "1968-03-01", "1968-04-01", "1968-02-01",
"1968-03-01", "1968-04-01"), value = c(14.1, 9.5, 22.1, 15.1,
13.5, 23.1, 16.1, 15.5, 13.1), uname = c(NA, NA, NA, NA, NA,
NA, NA, NA, NA), tcode = c(NA, NA, NA, NA, NA, NA, NA, NA, NA
)), row.names = c(NA, -9L), class = c("tbl_df", "tbl", "data.frame"
))