r: nesting a for loop within a custom function

Question

I have an anonymous function that I use to apply a double-exponential transformation to, and it works well for me with hard-coded values:

custom_func <- function(x) {
  0.01^(0.95^(x/max(x)*100))
}

df$newVar <- custom_func(df$var)

df$newVar

It returns the expected transformed variable:

However, I want to use create a version that will ingest multiple parameters for the second exponent and add them as permanent vectors to my data frame. My attempt doesn't add anything to the dataframe and having trouble understanding how to fix this:

alpha <- seq(0.85, 0.95, by= .01)

dblExponential <- function(a,x){
  for (i in alpha) {
    0.01^(a^(x/max(x)*100))
  }
}

dblExponential(alpha, df$var)

df

ah bon · Accepted Answer · 2022-01-23 15:13:15Z

1

Functions in R do not need an explicit return statement if you only want to use the value of the last evaluated statement. That works nicely in your simple function (your first code block).

In the second case, if I understand correctly, you want to return multiple time series that are computed one after the other in a loop. That is, you will have to compose the result data frame explicitly in the function and then return this result data frame.

An example based on your second code block

# Function definition (Note that the common convention is to define functions at the
#        top of any script so that people can understand the structure more easily.)
#
dblExponential<- function(a, x){
  # Prepare result data frame
  result_df <- data.frame(x)
  # loop through the sequence
  for (i in a) {
    # Compute column for value i
    result_column <- 0.01^(i^(x/max(x)*100))
    # Define name of column in the resulting data frame
    result_column_name <- as.character(i)
    # Add the column to the data frame
    result_df[result_column_name] <- result_column
  }
  # Return the result data frame
  return(result_df)
}

# Create example data frame
df <- data.frame(c(1, 2, 3, 4, 5, 6, 7))
colnames(df) <- c("var")

# Your sequence of exponents
alpha <- seq(0.85, 0.95, by= .01)

# Call to the function, assign the return value to df2
df2 <- dblExponential(alpha,df$var)

# print df2
df2

edited Jan 23, 2022 at 15:13

ah bon

10.1k22 gold badges82 silver badges185 bronze badges

answered Nov 23, 2018 at 13:37

0range

2,1762 gold badges25 silver badges32 bronze badges

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

2 Comments

dre Over a year ago

Excellent, that worked thank you. Two follow-ups: 1. Now all i need to do to add it to a new dataframe with the rest of my data is run cbind on this and the old dataframe? 2. None of the info on this site or the documentation explain the 'order of operations' or how to structure a function this complex. Any recommendations on what to read? Thanks again!

0range Over a year ago

@enriqueStateSpacias The function above gives you a new dataframe that includes the original vector. If you want to just include the new columns into the original dataframe df, you could 1) use cbind to combine df and df2, 2) modify df directly in the function (it's a global variable) but that would be bad style, 3) hand the entire dataframe over to the function, not just the column, or 4) put the function call in a loop instead of including the loop in the function. I am not sure I understand your second question. You may want to read about functional programming in general.

Collectives™ on Stack Overflow

r: nesting a for loop within a custom function

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related