CODEWITHSUNDEEP
r
Nora Elvestad
Nora Elvestad

Reputation: 1

how to change a single value based on properties of another in a dataframe in R?

I am trying to alter the value of a dummy variable for one country for three consecutive years. I have tries doing the following Try 1;

d[14696, "abort"] = 1

Try 2:

d$abort <- ifelse(d$drap == 7.826785, 1, d$abort)

d$abort <- ifelse(d$drap == 8.263874, 1, d$abort)

d$abort <- ifelse(d$drap == 12.060453, 1, d$abort)

Try 3:

d[14698, 6] = 1

Try 4:

d$abort[14698]  <-  1

*the number 14698 is the rownumber

None of them gives any errors which makes me assume they work. However, nothing changes in the data set.. Can someone help?

This is what the data frame looks like for Uruguay which needs to be edited for 2016-2019 on the variable abort:

Year Country Unemp Poverty GDPpc abort drap
2012 Uruguay 8.8379 0.2 12463 o 6.425273
2013 Uruguay 8.3742 0.2 10849 1 5.426374
2014 Uruguay 8.1894 0.2 12938 1 6.738383
2015 Uruguay 8.2349 0.3 11823 1 7.377327
2016 Uruguay 9.2369 0.3 10828 o 7.826785
2017 Uruguay 9.6273 0.4 10748 o 8.263874
2018 Uruguay 8.2374 0.3 11038 o 12.060453

Upvotes: 0

Views: 384

Answers (3)

Nora Elvestad
Nora Elvestad

Reputation: 1

I found out through the help of my professor that i had to load my dataset in a different way:

d <- d[,c("year", "cname", "wdi_unempilo", "wdi_povgap190", "mad_gdppc", "cai_request", "wdi_homicides")]

after that i had to create a new variable for each year, then use this to recode the original variable which was causing issues, and then delete the created variables:

create new variables

d$abort_2016 <- ifelse(d$Country=="Uruguay" & d$Year==2016,  1, d$abort)

d$abort_2017 <- ifelse(d$Country=="Uruguay" & d$Year==2017,  1, d$abort)

d$abort_2018 <- ifelse(d$Country=="Uruguay" & d$Year==2018,  1, d$abort)

use created variables to recode

d$abort <- ifelse(d$abort_2016 == 1, 1, d$abort)

d$abort <- ifelse(d$abort_2017 == 1, 1, d$abort)

d$abort <- ifelse(d$abort_2018 == 1, 1, d$abort)

Subset to remove created variables

d <-  subset(d, select = -c(abort_2016,abort_2016) )

d <-  subset(d, select = -c(abort_2017,abort_2017) )

d <-  subset(d, select = -c(abort_2018,abort_2018) )

This solved my issue

Upvotes: 0

Chris
Chris

Reputation: 472

Adding to the answer of rjen (while also using his/her dataframe df ) I would suggest you might use dplyr::case_when, such that

library(dplyr)


df %>% mutate(Abort = 
                case_when(drap %in% c(7.826785, 8.263874, 12.060453) ~ 1  ## if true
                          ,TRUE ~ Abort) ## if false condition
)

Here, if drap takes a certain value (similary to your 2nd approach) Abort becomes 1. Otherwise Abort stays the same. It might be more useful than the regular ifelse condition, as you can connect even more conditions together. You could additionally add something like

 case_when(drap %in% c(7.826785, 8.263874, 12.060453) & Year == 2013 ~ 1

Upvotes: 0

rjen
rjen

Reputation: 1972

The below is a tidyverse option based on the assumptions that drap is of type double and that x. is really NA.

library(dplyr)

mutate(df, Abort = if_else(!is.na(drap), 1, Abort))

## A tibble: 7 x 4
#    Year Country Abort  drap
#    <dbl>  <chr> <dbl> <dbl>
# 1  2012 Uruguay     0 NA   
# 2  2013 Uruguay     1 NA   
# 3  2014 Uruguay     1 NA   
# 4  2015 Uruguay     1 NA   
# 5  2016 Uruguay     1  7.83
# 6  2017 Uruguay     1  8.26
# 7  2018 Uruguay     1 12.1

Data

df <- structure(list(Year = c(2012, 2013, 2014, 2015, 2016, 2017, 2018
), Country = c("Uruguay", "Uruguay", "Uruguay", "Uruguay", "Uruguay", 
"Uruguay", "Uruguay"), Abort = c(0, 1, 1, 1, 0, 0, 0), drap = c(NA, 
NA, NA, NA, 7.826785, 8.263874, 12.060453)), row.names = c(NA, 
-7L), class = c("tbl_df", "tbl", "data.frame"))

Upvotes: 1

Related Questions