CODEWITHSUNDEEP
rif-statement
Lyndon L.
Lyndon L.

Reputation: 39

Ifelse statement syntax multiple conditions is broken

I am having some trouble with a nested ifelse statement. I have the following code written. It should force a TRUE or FALSE but I am getting some NAs. Can you help me figure out where its broken?

data looks something like...

Name<- c(Jon,Jim,Jake,Jess,Jill,Jay,Jason)
LinkFlag<- c( NA,  NA,  NA,  NA,  NA, NA, NA)
1<- c(1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0)
2<- c(2.5,  NA,  NA,  NA,  NA, 3.0,  NA)

spread_wide<-c(Name,LinkFlag,1,2)

The intended Result is that there are no NAs

sw_RESULTS$EvalSeq1to2<-ifelse(spread_wide$`1` == 99, TRUE, 
                   ifelse(spread_wide$`1` == (spread_wide$`2`-1), TRUE,  
                           ifelse(is.na(spread_wide$`2`),TRUE,   
                                       ifelse(spread_wide$`2` ==99,TRUE,
                                            ifelse(spread_wide$`1`== (spread_wide$`2`-0.5),TRUE, 
                                                       ifelse(spread_wide$`1` == spread_wide$`2`, TRUE,  
                                                                   ifelse(spread_wide$`LinkFlag`=='FlagforSkip", TRUE,
                                                                                                                                 FALSE)))))))

Upvotes: 1

Views: 192

Answers (1)

thelatemail
thelatemail

Reputation: 93813

You essentially have a set of binary checks that each return TRUE, or FALSE if none of them are true. So, it's simple then to just make a matrix or list of all the checks together:

checks <- with(spread_wide,
  cbind(
    `1` == 99,
    `1` == `2` - 1,
    is.na(`2`),
    `2` == 99,
    `1` == (`2` - 0.5),
    `1` == `2`,
    LinkFlag == 'FlagforSkip'
  )
)

rowSums(checks, na.rm=TRUE) > 0
#[1] FALSE  TRUE  TRUE  TRUE  TRUE FALSE  TRUE

The secret to the final line is using the na.rm=TRUE so that you are never giving any consideration to comparisons like NA == 1 which will return NA instead of TRUE or FALSE.

As a side-note, using variable names like 1 or 2 is just asking for trouble.

Reproducible data that works:

Name<- c("Jon","Jim","Jake","Jess","Jill","Jay","Jason")
LinkFlag<- c( NA,  NA,  NA,  NA,  NA, NA, NA)
`1` <- c(1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0)
`2` <- c(2.5,  NA,  NA,  NA,  NA, 3.0,  NA)
spread_wide<- data.frame(Name,LinkFlag,`1`,`2`,check.names=FALSE)

Upvotes: 2

Related Questions