How to create new function from output of previous function in R?

Question

I am ignorant when it comes to R programming and programming in general but I have two pieces of code that have come across a similar problem (for me). Here we go...

(A)

I currently have a function that returns record(s) of a patient, trial number, and other information. It looks like this:

     ID trial     start   finish     mark     mean    number
903 A34    19     90910 18775077     8236  -0.0197  1.972876
904 A34    19  18782377 23089165     2343   0.0374  2.052525
905 A34    19  23093018 43203507    10267  -0.0162  1.977668
906 A34    19  43203990 43447468       93   0.2138  2.319478
907 A34    19  43447802 43663369      112  -0.0355  1.951387
908 A34    19  43663624 43834506       80  -0.5385  1.376973
909 A34    19  43834848 59097854     8655  -0.0095  1.986873

Below is the code I have written for it.

getRS <- function(CNA, samples = NULL, trial = NULL){ race <- racing.summary(subset(CNA, samplelist = samples, triallist = trial)) race$number <- (2^race$mean)*2 return(race) }

I am wondering if it is possible to use this output in a new function to do simple arithmetics. I am looking to subtract 'finish' from 'start' to create 'length', create a new 'mean' with all the means from above and extract the largest 'number' to create 'max.number' whilst not displaying 'mark' at all.

An output similar to this:

ID    trial     max.length          mean    max.number
A34       19       20110489   -0.05260000     2.3194777

AND/OR

(B)

I have an alternative function that creates a data frame of ALL the patients with the already calculated data. I used this code:

getSum <- function (){
  race_mean <- as.data.frame(df %>% group_by(ID, trial) %>% summarise(mean = mean(mean)))
  race_length <- as.data.frame(df %>% group_by(ID,trial) %>% summarise(max.length = max(end - start)))
  seg_number <- as.data.frame(df %>% group_by(ID,trial) %>% summarise(max.number = max(number)))
  race_m_l_merge <- as.data.frame(merge(x = race_length, y = race_mean))
  race_m_l_n_merge <- as.data.frame(merge(x = race_m_l_merge, y = race_number))
  ordered_summary <- as.data.frame(race_m_l_n_merge[order(race_m_l_n_merge$trial),])
  View(ordered_summary)
}

Which gives an output like this:

      ID trial    max.length         mean       max.number
1    A22     1      96637812   -1.648909e-01     2.6989533
25   A23     1     101363101   -6.275455e-02     2.2468441
49   A24     1      72598875   -5.878000e-02     2.8204004
73   A25     1     112628591   -3.346917e-01     2.0675182
97   A26     1      55490417    7.621429e-02     2.4512200
121  A28     1     130879821   -4.218571e-02     2.0679481
145  A29     1      72590096   -3.093417e-01     2.3450196
169  A30     1      32642030    4.242500e-02     2.6375528
193  A32     1      34350731   -8.188372e-02     2.1149155
217  A33     1      77537981   -1.305833e-01     2.1125713

With this, I would like to create a function as to specify which ID and which trial I would like to lookup like so: Function("A22",1) .

I'm hoping that my R Script for the future would work arbitrarily for future endeavors so any help would be much appreciated either on my question A, B or perhaps both! Or even suggestions for links to helpful websites. :)

Answer 1

If you have already defined your functions getRS and getSum , then you can call them inside a new function.

Uou just have to change the line that contains View(ordered_summary) in getSum to return(ordered_summary) or simply ordered_summary , so you it returns an object you can further manipulate.

lookup_function <- function(data_lookup, id_lookup, trial_lookup) {
  data_df <- getRS(CNA = data_lookup)
  summary_df <- getSum(df = data_df)
  subset(x = results_df, subset = (ID == id_lookup & trial == trial_lookup))
}

You can write this function in a concise way, if you feel inclined to do so.

lookup_function <- function(data_lookup, id_lookup, trial_lookup) {
  subset(x = getSum(getRS(data_lookup)), subset = (ID == id_lookup & trial == trial_lookup))
}

Or, if you don't want to have three different functions, you can create a function that has getRS and getSum defined inside itself.

lookup_function <- function(data_lookup, id_lookup, trial_lookup) {
  data_df <- getRS(CNA = data_lookup)
  summary_df <- getSum(df = data_df)
  subset(x = results_df, subset = (ID == id_lookup & trial == trial_lookup))
}

lookup_function <- function(data_lookup, id_lookup, trial_lookup) {
  getRS <- function(CNA, samples = NULL, trial = NULL){
    race <- 
      racing.summary(subset(CNA, samplelist = samples, triallist = trial))
    race$number <- 
      (2 ^ race$mean) * 2

    race
  }

  getSum <- function(df) {
    unordered_summary <- 
      df %>% 
      group_by(ID, trial) %>% 
      summarise(mean = mean(mean),
                max.length = max(end - start),
                max.number = max(number)) %>% 
      data.frame()

    ordered_summary <- 
      data.frame(unordered_summary[order(unordered_summary$trial), ])

    ordered_summary
  }

  data_df <- getRS(CNA = data_lookup)

  summary_df <- getSum(df = data_df)

  subset(x = results_df, subset = (ID == id_lookup & trial == trial_lookup))
}

I have edited the code for getSum , as I didn't see a reason to call summarize three times, instead of a single time. You can use your own function, of course, as I don't know the particulars of your task at hand.