2.6 Functions and Loops

Components of a basic function

my_function <-function(x,y){ # create function as object (my_function <- function()) 
                             # define inputs of function (function(x,y))
                             # function is written between {} brackets
result=x^2+y^2               # what your function does with inputs
return(result)               # return() gives the result as the output
}
                         
my_function(2,3)              # now try the function out

## [1] 13

A function with conditions

largeCpGi<-function(bedRow){      # function takes input one row of CpGi data frame 
                                  # generated earlier under "Read files in base R"
 cpglen=bedRow[3]-bedRow[2]+1     # subtract column 3 value from column 2 value + 1
 if(cpglen>1500){                 # if the result (cpglen) meets this parameter                                         # (is larger than 1500)
    cat("this is large\n")        # cat[used in place of print/return] "this is                                         # large\n" --> the \n means print a new line
 }
 else if(cpglen<=1500 & cpglen>700){           # else if -- if the above was not true,                                                # see if result meets these parameters
    cat("this is normal\n")
 }
 else{                            # if the result did not meet previous parameters
    cat("this is short\n")        
 }
}
largeCpGi(cpgi.df[10,])           # function to take 10th row of data frame 

## this is short

largeCpGi(cpgi.df[10])            # what happens if you input this instead?

## Error in `[.data.frame`(cpgi.df, 10): undefined columns selected

largeCpGi(cpgi.df(10))            # or this?

## Error in cpgi.df(10): could not find function "cpgi.df"

Loops to repeat a function multiple times

result=c() # this is where we will keep the lengths. for now it is an empty vector.

for(i in 1:100){                      # start a loop to iterate 100 times
    len=cpgi.df[i,3]-cpgi.df[i,2]+1   # to calculate the length of each cpgi
                                      # where i is each row of the data frame cpgi.df
    result=c(result,len)              # append the length to the result each time
}

result                                # check the result

##   [1]  277  360  383 1325 3498 1602 1317  250  871  226  788  300 1001  311  790
##  [16]  224 1259  725  525  690  426  636  827 1010  237  601  484  297  968  441
##  [31]  793  235  212  648  656  659  794  324  356  729  786  202  401 1787 1045
##  [46]  265  203  498  455  203  821  745 1013  341 1795  801  333  885  520  497
##  [61] 1403 1737 1192 1007 1142 1009  276  626 1440  512  552  344  268  233  273
##  [76]  232 2353  456  937  941  985 1584  878  357  758  884 1355  979 1294  757
##  [91]  786  807 1118  833  931  533  618 1965  321 1048

Loops are not always the most efficient way to work in R. We can use special functions to perform loops.

In base R: the apply family lapply applies a function to a single list or vector mapply is like lapply but over multiple vectors/lists

vectorized function can be summed (+, rowSums, colSums)

In the tidyverse: purrr –> specifically map