[R] problem with function that adds rows to dataframe based on conditional statement

[Tom Wright]

If all you want is to add a row of na's could you just do something
like:

nExpectedRows<-length(unique(animals)) * length(unique(animalYears)) * 2

newDf<-data.frame(animals=rep(NA,nExpectedRows-nrow(comAn)),
		  animalYears=rep(NA,nExpectedRows-nrow(comAn)),
		  animalMass=rep(NA,nExpectedRows-nrow(comAn)))

comAn = rbind(comAn,newDf)



On Thu, 2015-03-05 at 13:41 -0700, Curtis Burkhalter wrote:
> Hello everyone,
> 
> I'm having a problem with a function that I wrote that is supposed to add a
> row to dataframe based upon a conditional statement. To explain I've used
> an example below:
> 
> #create data frame
> animals=c("bird","dog","cat")
> animals=rep(animals,each=4)
> animals=animals[1:11]
> animalYears=c(1,1,2,2,1,1,2,2,1,1,2)
> animalMass=round(runif(11,min=10,max=50),0)
> 
> comAn=as.data.frame(cbind(animals,animalYears,animalMass))
> comAn
> 
>   * animals* *animalYears* *animalMass*
> 1     bird           1         30
> 2     bird           1         32
> 3     bird           2         27
> 4     bird           2         16
> 5      dog           1         22
> 6      dog           1         25
> 7      dog           2         41
> 8      dog           2         22
> 9      cat           1         30
> 10     cat           1         37
> 11     cat           2         49
> 
> We can see here that for every type of animal I have two years of mass
> measurements, except for the cat in year 2. What I want to do is add an
> additional row to the end of the dataframe that consists strictly of NAs
> and then I can substitute those out later.
> 
> So what I first did was split the 'comAn' dataframe into the different
> Animal by Year combos.
> 
> #This line splits 'com_An' into a list ordered by the Animal by Year combos
> comAn_split=split(comAn, paste(comAn$animals,comAn$animalYear))
> 
> Then I wrote the function that identifies whether a particular Animal by
> Year combo is less than two rows in length and if so it should add another
> row that consists only of NAs using the vector 'NAs':
> 
> #This function identifies the length of each Animal by Year combo and then
> #uses the rbind function built in R to add a row
> #to each animal by year combo if they have less than 2 samples
> 
> addNA <- function(comAn) {
>   NAs=c(NA,NA,NA)
>         ind <- seq_len(nrow(comAn))
>         comAn[ifelse(length(ind)<2,rbind(NAs),length(ind)),]
> }
> 
> #This applies the function addNs to the animals data organized in list
> format
> addedNAcomAn <- do.call(rbind, lapply(comAn_split, addNA))
> addedNAcomAn
> 
> When I apply the function to the list of the different Animal by Year
> combos this is what I get:
>        animals animalYears animalMass
> bird 1    bird           1         23
> bird 2    bird           2         50
> cat 1      cat           1         15
> cat 2     <NA>        <NA>       <NA>
> dog 1      dog           1         23
> dog 2      dog           2         38
> 
> What I expect is this:
> 
>    animals animalYears animalMass
> 1     bird           1         41
> 2     bird           1         23
> 3     bird           2         23
> 4     bird           2         50
> 5      dog           1         49
> 6      dog           1         23
> 7      dog           2         13
> 8      dog           2         38
> 9      cat           1         42
> 10     cat           1         15
> 11     cat           2         33
> 12     NA           NA      NA
> 
> Am I conditioning improperly within the function or am I missing something
> else. Any help would be greatly appreciated.
> 
> Best
>