In R Convert to Date of several columns?

Question

Can you give me a hand with the code below? I did try to find an answer to this but migth have missed, if there is one already sorry for your time.

I have a DataFrame like the exemple below. What I need to do is to convert all dt_ variables to date. I got it with mutate()/lapply one by one but I was looking for an automatic method. I am working in R.

co_cid Tipo  dt_notificacao co_uf_notificac~ co_uf_completo no_municipio_no~ dt_diagnostico_~
  <fct>  <fct> <fct>          <fct>            <fct>          <fct>            <fct>           
1 A90    Deng~ 01/10/2016     PE               PERNAMBUCO     Recife           23/01/2015      
2 A90    Deng~ 02/11/2016     PE               PERNAMBUCO     Recife           09/01/2015      
3 A90    Deng~ 01/11/2016     PE               PERNAMBUCO     Recife           12/12/2015      
4 A90    Deng~ 02/04/2016     PE               PERNAMBUCO     Recife           12/08/2015      
5 A90    Deng~ 01/08/2016     PE               PERNAMBUCO     Recife           12/01/2015      
6 A90    Deng~ 01/11/2016     PE               PERNAMBUCO     Recife           12/04/2015  

I got all the dt_, that should be dates with:

dt_vec <- nomes_colunas[(sapply(nomes_colunas, startsWith,prefix = "dt_"))]

Then I wanted to use it to convert all dt_ columns to Date and replace it in the original df. I tryed a for loop with mutate but the column name stays with the variable name. So I end up with only one 'coluna' variable with the Dates.

for (coluna in dt_vec) {
  df_dados <- df_dados %>% mutate(coluna, coluna = as.Date(coluna, format = "%d/%m/%Y"))
}

As for apply I found it hard to replace the values in the original df.

Thanks in advance!

Answer 1

If we want to convert several columns, use mutate with across

library(lubridate)
library(dplyr)#1.0.0  
df_dados <- df_dados %>%
                mutate(across(starts_with('dt_', dmy)))

In the earlier versions of dplyr , mutate_at can be used

df_dados <- df_dados %>%
              mutate_at(vars(starts_with('dt_')), dmy)

Answer 2

You can find out the columns to transform using startsWith and use lapply to convert them to dates using as.Date . We can do this using base R.

cols <- startsWith(names(df), 'dt')
df[cols] <- lapply(df[cols], as.Date, "%d/%m/%Y")
df

#  co_cid  Tipo dt_notificacao co_uf_notificac co_uf_completo no_municipio_no dt_diagnostico_
#1    A90 Deng~     2016-10-01              PE     PERNAMBUCO          Recife      2015-01-23
#2    A90 Deng~     2016-11-02              PE     PERNAMBUCO          Recife      2015-01-09
#3    A90 Deng~     2016-11-01              PE     PERNAMBUCO          Recife      2015-12-12
#4    A90 Deng~     2016-04-02              PE     PERNAMBUCO          Recife      2015-08-12
#5    A90 Deng~     2016-08-01              PE     PERNAMBUCO          Recife      2015-01-12
#6    A90 Deng~     2016-11-01              PE     PERNAMBUCO          Recife      2015-04-12