common_functions.Rmd

---
  Introduction to several commonly used functions in R
--

## with()
The with( ) function applys an expression to a dataset. It is similar to 
DATA= in SAS.

```{r}
# with(data, expression)
# example applying a t-test to a data frame mydata 
with(mydata, t.test(y ~ group))

```
Please look at other examples [here](http://stackoverflow.com/questions/21827572/what-is-the-difference-between-with-and-within-in-r) and [here](http://www.magesblog.com/2012/01/say-it-in-r-with-by-apply-and-friends.html).

## by()
The by( ) function applys a function to each level of a factor or factors. 
It is similar to BY processing in SAS.

```{r}
# by(data, factorlist, function)
# example obtain variable means separately for
# each level of byvar in data frame mydata 
by(mydata, mydatat$byvar, function(x) mean(x))
```

Please look [here](http://www.magesblog.com/2012/01/say-it-in-r-with-by-apply-and-friends.html) for more details.

## do.call()
do.call calls a function with a list of arguments, lapply applies a function to each element of the list

```{r}
do.call(sum, list(c(1,2,4,1,2), na.rm = TRUE))
#10
lapply(c(1,2,4,1,2), function(x) x + 1)
#2
#3
#5
#2
#3
do.call("+",list(4,5))
#9
```
More examples [here](https://stat.ethz.ch/R-manual/R-devel/library/base/html/do.call.html).

## more()
more() is a user-defined function that is helpful in printing out a large object. Taken from [here](http://stackoverflow.com/questions/21610468/pause-print-of-large-object-when-it-fills-the-screen).

```{r}
# to print out an object such as data.frame mydf 20 lines at a time, use:
more(mydf)

# where more() is defined as

more <- function(expr, lines=20) {
  out <- capture.output(expr)
  n <- length(out)
  i <- 1
  while( i < n ) {
    j <- 0
    while( j < lines && i <= n ) {
      cat(out[i],"\n")
      j <- j + 1
      i <- i + 1
    }
    if(i<n){
      rl <- readline()
      if( grepl('^ *q', rl, ignore.case=TRUE) ) i <- n
      if( grepl('^ *t', rl, ignore.case=TRUE) ) i <- n - lines + 1
      if( grepl('^ *[0-9]', rl) ) i <- as.numeric(rl)/10*n + 1
    }
  }
  invisible(out)
}
```

## options()
options() can be used to increase the limit for max.print in R. More info [here](http://stackoverflow.com/questions/6758727/how-to-increase-the-limit-for-max-print-in-r).

```r
options(max.print=1000000)
```

## To check which columns in the data frame `df` have missing values
```r
colnames(df)[colSums(is.na(df)) > 0]
```