Creating variables names using objects in a for loop in R

Question

I think this might be a very simple question, but I can't for the life of me find the answer online or in the book I've been using to learn R.

I'm trying to create a table with variables named based on the values in a vector of an existing matrix. Here is an example of how the vectors or interest appear in the table (named "thresholds") I am using where the variable names appears-

varname       threshold_1_name threshold_1_value
1   BMXBMI    high             25
2  BMXCALF    low              40
3    BMXHT    high             180 
4   BMXLEG    low              40   
5   BMXSUB    high             32  
6 BMXTHICR    high             65

The table has 81 records in it, and I want to do something like this:

for (i in 1:81) {
  varname1 <- paste(thresholds$varname[i], thresholds$threshold_1_name[i], sep = "_")
  newtable$[varname1] <- ifelse((bigTable$[thresholds$varname[i]] < thresholds$threshold_1_value[i]),1,0)
}

which would create 'newtable' with 81 columns with names where the first six columns would be named BMXBMI_high, BMXCALF_low, BMXHT_high, BMXLEG_low, BMXSUB_high, BMXTHICR_high. My ifelse statement seems to be fine- I tested it outside of the loop and it worked. I think I'm using incorrect syntax to create the variable names.

Any advice on what I should do or how I should search for an answer would be greatly appreciated. I think part of my inability to find an answer is because I'm using incorrect vocabulary/search terms. Thanks!

@Ben- as you seem to have predicted, I'm not having issues with my ifelse line. Here is a sample of bigTable (and I added a column to the 'threshold' sample above) to help you provide some advice on how to sort the issue there. I am trying to code values in the new variables as 0/1 depending on whether the value in bigTable is above or below the value in 'thresholds'

Sample of bigTable:

  BMXHT BMXBMI BMXLEG BMXCALF BMXWAIST BMXTHICR BMXTRI BMXSUB
1 174.0  24.90     NA    37.5     98.0       NA   12.8   20.4
2 178.3  29.10   45.2    42.6     99.9     56.2   17.4   38.6
3 162.0  22.56   39.7    34.0     81.6     47.0   20.3   16.8
4 162.9  29.39   43.0    37.2     90.7     55.7   26.4   34.2
5 190.1  30.94   46.6    43.7    108.0     64.0   15.5   26.6
6 180.0  30.62   46.0    40.5    112.8     57.1   26.2   NA

When I tried to code everything in one line, I keep getting an error that says the code is only reading the first entry, so I am now trying the following segment of code which is both horribly inefficient and still not working (the first two lines are what you previously sent)-

varname1 <- paste(thresholds$varname, thresholds$threshold_1_name, sep = "_")
bigTable[varname1[1:5]] <- NA

for (i in 1:5) {
  value <- thresholds$threshold_1_value[i]
  var <- thresholds$varname[i]
  newvar <- varname1[i]
  for(j in 1:10) {
    if(bigTable[var[j]] > value) {bigTable[newvar] = 1}
    else if (bigTable[var[j]] <= value) {bigTable[newvar] = 0}
  }
}

Again, any help you can provide is greatly appreciated!

How about varname1[i] and newtable$varname1[i] instead of varname1 and newtable$[varname1] on the LHS of the assignment in the loop? — Ben
– Ben, Commented Dec 20, 2012 at 22:35
Can you post bigTable as well. There are far quicker ways to do this, but you also need to understand the difference between $ and [[. Look at ?Extract. — mnel
– mnel, Commented Dec 20, 2012 at 22:38

Ben · Accepted Answer · 2012-12-21 08:41:42Z

2

Here's an answer to the question of 'how do I add many columns to a dataframe using variables in the dataframe', which seems to be part of the original problem (can't do much about the rest until we see what bigTable looks like):

# prepare data
thresholds <- read.table(text = "varname       threshold_1_name 
   BMXBMI    high 
  BMXCALF    low
    BMXHT    high
   BMXLEG    low
   BMXSUB    high
 BMXTHICR    high", header = TRUE)

To pursue the loop in the question, we can use it create new column names based on existing data

varname1 <- NULL
for (i in 1:nrow(thresholds) {
  varname1[i] <- paste(thresholds$varname[i], thresholds$threshold_1_name[i], sep = "_")
}

But note that a loop isn't needed here, a basic vector operation will get the same result as the loop:

varname1 <- paste(thresholds$varname, thresholds$threshold_1_name, sep = "_")

Anyway, whichever way you do it, then you can add the names as column names like so:

# add new columns to a new dataframe
newtable <- data.frame(setNames(replicate(length(varname1), numeric(0), simplify = F), varname1))

And here's the output, new columns with names that are a function of existing variables:

 str(newtable)
'data.frame':   0 obs. of  6 variables:
 $ BMXBMI_high  : num 
 $ BMXCALF_low  : num 
 $ BMXHT_high   : num 
 $ BMXLEG_low   : num 
 $ BMXSUB_high  : num 
 $ BMXTHICR_high: num

edited Dec 21, 2012 at 8:41

answered Dec 20, 2012 at 22:51

Ben

42.6k21 gold badges142 silver badges232 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

Anthony Damico Over a year ago

the columns shouldn't be added to thresholds they should be added to a new table called newtable :)

Struggling_with_R Over a year ago

@Ben- I just edited the post above to include a sample of bigTable. Any help you can provide is greatly appreciated.

Ben Over a year ago

@Struggling_with_R, can you add to your question a description of what the loop involving bigTable is supposed to do? It's not really clear what you're after with that...

Collectives™ on Stack Overflow

Creating variables names using objects in a for loop in R

1 Answer 1

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related