R help - Oct 2020 - Error: variable not found

Hello

I have a question. I made an r-script and did a few commands needed to make some
new variables. They all work out well, and when I run the commands, the new
variables appear in the dataset. I can also work with these new variables to
make other new variables from them. Also, when I use summary(dataset), those new
variables appear in the summary of the dataset.
But, when I do summary(new variable) => error: variable not found.
For example: summary(couple_id) => Error in summary(couple_id) : object
'couple_id' not found. What can I do about this?

R-script:
attach(ipumsi_00008_dta)
library(tinytex)
library(dplyr)
library(ggplot2)
library(tidyr)
library(knitr)
library(forcats)
library(mice)
library(pander)
library(ggcorrplot)
library(lubridate)
# true/false code when sploc is greater than zero and sprule is equal to 1 or 2
ipumsi_00008_dta <- mutate(ipumsi_00008_dta, rule_union = sploc>0 &
(sprule==1 | sprule==2))
## creating numeric code for rule_union & rule_unionn: 1 when sploc is
greater than zero and sprule is equal to 1 or 2, 0 if not.
## This is neccesary because otherwise it is a logical code and we cannot
multiply with it, which is needed
ipumsi_00008_dta <- ipumsi_00008_dta %>% mutate(rule_union_numeric =
as.numeric(rule_union))
### creating unique numeric code for sploc / pernum variables
ipumsi_00008_dta <- mutate(ipumsi_00008_dta, sploc_pernum_code =
pernum*(sex==1) + sploc*(sex==2))
#### dividing serial by 1000, otherwise, the ultimate couple_id is too large,
and it works in this dataset because the serials start at 1000 (I will have to
check if this works for other datasets)
ipumsi_00008_dta <- mutate(ipumsi_00008_dta, serial_divided = serial%/%1000)
##### creating unique union identifier
ipumsi_00008_dta$union_id = paste0(ipumsi_00008_dta$country,
ipumsi_00008_dta$year, ipumsi_00008_dta$serial_divided,
ipumsi_00008_dta$sploc_pernum_code)
ipumsi_00008_dta <- ipumsi_00008_dta %>% mutate(union_id_numeric =
as.numeric(union_id))
ipumsi_00008_dta <- mutate(ipumsi_00008_dta, couple_id =
union_id_numeric*rule_union_numeric)
###### Understanding the couple_id variable
summary(couple_id)
summary(ipumsi_00008_dta)

I also attached my dataset.

Thank you very much for the answer!
Hannah Van Impe

Hi Hannah,
Have you tried:

summary(ipumsi_00008_dta$couple_id)

Jim

On Fri, Oct 30, 2020 at 7:34 PM Hannah Van Impe <hannahvanimpe at
outlook.com>
wrote:
> Hello
>
> I have a question. I made an r-script and did a few commands needed to
> make some new variables. They all work out well, and when I run the
> commands, the new variables appear in the dataset. I can also work with
> these new variables to make other new variables from them. Also, when I use
> summary(dataset), those new variables appear in the summary of the dataset.
> But, when I do summary(new variable) => error: variable not found.
> For example: summary(couple_id) => Error in summary(couple_id) : object
> 'couple_id' not found. What can I do about this?
>
> R-script:
> attach(ipumsi_00008_dta)
> library(tinytex)
> library(dplyr)
> library(ggplot2)
> library(tidyr)
> library(knitr)
> library(forcats)
> library(mice)
> library(pander)
> library(ggcorrplot)
> library(lubridate)
> # true/false code when sploc is greater than zero and sprule is equal to 1
> or 2
> ipumsi_00008_dta <- mutate(ipumsi_00008_dta, rule_union = sploc>0
&
> (sprule==1 | sprule==2))
> ## creating numeric code for rule_union & rule_unionn: 1 when sploc is
> greater than zero and sprule is equal to 1 or 2, 0 if not.
> ## This is neccesary because otherwise it is a logical code and we cannot
> multiply with it, which is needed
> ipumsi_00008_dta <- ipumsi_00008_dta %>% mutate(rule_union_numeric
> as.numeric(rule_union))
> ### creating unique numeric code for sploc / pernum variables
> ipumsi_00008_dta <- mutate(ipumsi_00008_dta, sploc_pernum_code >
pernum*(sex==1) + sploc*(sex==2))
> #### dividing serial by 1000, otherwise, the ultimate couple_id is too
> large, and it works in this dataset because the serials start at 1000 (I
> will have to check if this works for other datasets)
> ipumsi_00008_dta <- mutate(ipumsi_00008_dta, serial_divided >
serial%/%1000)
> ##### creating unique union identifier
> ipumsi_00008_dta$union_id = paste0(ipumsi_00008_dta$country,
> ipumsi_00008_dta$year, ipumsi_00008_dta$serial_divided,
> ipumsi_00008_dta$sploc_pernum_code)
> ipumsi_00008_dta <- ipumsi_00008_dta %>% mutate(union_id_numeric >
as.numeric(union_id))
> ipumsi_00008_dta <- mutate(ipumsi_00008_dta, couple_id >
union_id_numeric*rule_union_numeric)
> ###### Understanding the couple_id variable
> summary(couple_id)
> summary(ipumsi_00008_dta)
>
> I also attached my dataset.
>
> Thank you very much for the answer!
> Hannah Van Impe
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
	[[alternative HTML version deleted]]

Dear Hannah

I think the problem is that attach() is not doing what you think it is. 
It does seem to make it easy to make mistakes. I would suggest switching 
to using with() instead or using the data = parameter to functions which 
support it.

Michael

On 30/10/2020 08:15, Hannah Van Impe wrote:
> Hello
> 
> I have a question. I made an r-script and did a few commands needed to make
some new variables. They all work out well, and when I run the commands, the new
variables appear in the dataset. I can also work with these new variables to
make other new variables from them. Also, when I use summary(dataset), those new
variables appear in the summary of the dataset.
> But, when I do summary(new variable) => error: variable not found.
> For example: summary(couple_id) => Error in summary(couple_id) : object
'couple_id' not found. What can I do about this?
> 
> R-script:
> attach(ipumsi_00008_dta)
> library(tinytex)
> library(dplyr)
> library(ggplot2)
> library(tidyr)
> library(knitr)
> library(forcats)
> library(mice)
> library(pander)
> library(ggcorrplot)
> library(lubridate)
> # true/false code when sploc is greater than zero and sprule is equal to 1
or 2
> ipumsi_00008_dta <- mutate(ipumsi_00008_dta, rule_union = sploc>0
& (sprule==1 | sprule==2))
> ## creating numeric code for rule_union & rule_unionn: 1 when sploc is
greater than zero and sprule is equal to 1 or 2, 0 if not.
> ## This is neccesary because otherwise it is a logical code and we cannot
multiply with it, which is needed
> ipumsi_00008_dta <- ipumsi_00008_dta %>% mutate(rule_union_numeric =
as.numeric(rule_union))
> ### creating unique numeric code for sploc / pernum variables
> ipumsi_00008_dta <- mutate(ipumsi_00008_dta, sploc_pernum_code =
pernum*(sex==1) + sploc*(sex==2))
> #### dividing serial by 1000, otherwise, the ultimate couple_id is too
large, and it works in this dataset because the serials start at 1000 (I will
have to check if this works for other datasets)
> ipumsi_00008_dta <- mutate(ipumsi_00008_dta, serial_divided =
serial%/%1000)
> ##### creating unique union identifier
> ipumsi_00008_dta$union_id = paste0(ipumsi_00008_dta$country,
ipumsi_00008_dta$year, ipumsi_00008_dta$serial_divided,
ipumsi_00008_dta$sploc_pernum_code)
> ipumsi_00008_dta <- ipumsi_00008_dta %>% mutate(union_id_numeric =
as.numeric(union_id))
> ipumsi_00008_dta <- mutate(ipumsi_00008_dta, couple_id =
union_id_numeric*rule_union_numeric)
> ###### Understanding the couple_id variable
> summary(couple_id)
> summary(ipumsi_00008_dta)
> 
> I also attached my dataset.
> 
> Thank you very much for the answer!
> Hannah Van Impe
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
> 
-- 
Michael
http://www.dewey.myzen.co.uk/home.html

Hi Hannah,
Using the same code I sent before, you can append the partner codes to the
household code. I apologize, but I don't know how to use the
dplyr/tidyr/... stuff so this is written in straight R code using logic
statements.

ipumsi_00008_dta<-
 read.table(
 text="country year sample serial hhwt pernum perwt resident sploc sprule
204 2013 204201301 4000 10 1 10 1 5 2
204 2013 204201301 4000 10 2 10 1 0 2
204 2013 204201301 4000 10 3 10 1 0 2
204 2013 204201301 4000 10 4 10 1 0 2
204 2013 204201301 4000 10 5 10 1 1 2
204 2013 204201301 4000 10 6 10 1 1 2
204 2013 204201301 4000 10 7 10 1 0 2
204 2013 204201301 4000 10 8 10 1 0 2
204 2013 204201301 4000 10 9 10 1 0 2
204 2013 204201301 7000 10 1 10 1 2 1
204 2013 204201301 7000 10 2 10 1 1 1
204 2013 204201301 7000 10 3 10 1 0 0
204 2013 204201301 7000 10 4 10 1 5 1
204 2013 204201301 7000 10 5 10 1 4 1",
header=TRUE,stringsAsFactors=FALSE)
for(hh in unique(ipumsi_00008_dta$serial)) {
 cat("hh",hh," ")
 for(ind in ipumsi_00008_dta$pernum[ipumsi_00008_dta$serial == hh]) {
  cat("ind",ind,"\n")
  if(ipumsi_00008_dta$sploc[ipumsi_00008_dta$serial == hh &
    ipumsi_00008_dta$pernum == ind] > 0) {
   cat("sploc > 0\n")
   relationships<-
    ipumsi_00008_dta$pernum[ipumsi_00008_dta$serial == hh &
     ipumsi_00008_dta$sploc == ind]
   cat(relationships,"\n")
   ipumsi_00008_dta$sprule[ipumsi_00008_dta$serial == hh &
    ipumsi_00008_dta$pernum == ind]<-
     paste(c(hh,relationships),sep="",collapse="_")
  } else {
   ipumsi_00008_dta$sprule[ipumsi_00008_dta$serial == hh &
    ipumsi_00008_dta$pernum == ind]<-hh
  }
 }
}
ipumsi_00008_dta

This appends the partner codes using "_" as a separator. You can do it
without the "_" and get a numeric variable, but I think you will
generate
ambiguous "sprule" codes. Again, this will not work with
between-household
relationships.

Jim

On Fri, Oct 30, 2020 at 11:16 PM Hannah Van Impe <hannahvanimpe at
outlook.com>
wrote:
> Thank you very much for the answer.
>
> I also have another question. With this data, I made the variable union_id
> using paste0. (I am writing a thesis and this part is necessary, but I
> don?t have previous knowledge of R, so it is difficult for me). My
> professor told me, that if I use paste0, it can be problematic because
> ?pernum? and ?serial? can have different numbers of digits. In other words,
> if pernum is 12 then paste0 prints 12, but if pernum is 1 paste0 prints 1
> instead of 01, so you will not have that the last two digits always
> correspond to pernum. He suggest me to use an alternative way for creating
> this union_id variable. Do you have any idea how I can do this?
>
>
	[[alternative HTML version deleted]]