R Programming Newbie! - Median Loop Function Broken - r

Question

So the function below is using a really big dataframe. Two columns of this DF are the year houses were built and the other is the cost.
I want to input the column name, in this case ds$Built as the function argument
YearCount <- the length of the number of unique values, years, in DS$Built
YearList <- a vector of the unique values, years, in ds$Built
Then I want it to do a for loop of YearCount number of iterations where it takes the median of values in Cost06 but only where the values in ds$Built = the value in YearList[i]
Take the resulting median and append it to the empty vector CostVec
At the end, make a 2 column dataframe of YearList and iMedian.
But this does not work. It is doing what you see below where it is assigning the same median value (the value from the last iteration in the loop) to all the years, rather than making each year's result available.
I'm a newbie so please thank you very much for your patience.
Median.DF <- function(x)
{
YearCount <- length(unique(x))
YearList <- unique(x)
CostVec <- c()
for (i in YearCount) {
imedian <- median(ds[x == YearList,"COST06"],na.rm = TRUE)
CostVec <- append(CostVec,imedian)
}
MedianCost.data <- data.frame(YearList, CostVec)
return(MedianCost.data)
}
YearList CostVec
1 2004 1629
2 2007 1629
3 2005 1629
4 1980 1629
5 1985 1629
6 2003 1629
7 2008 1629
8 1990 1629
9 1975 1629
10 1970 1629
11 1950 1629
12 1920 1629
13 1960 1629
14 1930 1629
15 1919 1629
16 1940 1629
17 1995 1629
18 2006 1629
19 2009 1629
20 2000 1629
21 2002 1629
22 2001 1629
23 2010 1629
24 2011 1629
25 2012 1629
26 2013 1629

Answer 1

Here is a better (cleaner, more R-like) way to do this. I'll leave it to you to turn it into the function.
df <- data.frame(YearList=sample(2000:2006,30, replace=T), CostVec=10^3+100*runif(30,0,1))
dfSplit <- split(df, df$YearList)
medianByYears <- lapply(dfSplit, function(x) median(x$CostVec))
medianByYearsClean <- do.call(rbind, medianByYears)
data.frame(Years=rownames(medianByYearsClean), MedianPrices=medianByYearsClean, row.names=NULL)
A few other tips:
Before you write a function, just test each line by line outside of a function. The main problem here is:
for (i in YearCount) { do stuff }
is analogous to
for (i in 5){print(i)}
Which does one thing once. You want to do:
for (i in seq(YearCount)) { do stuff }