Given a dataset, for example a CSV file that might look like this:
x,y
1,2
1,5
2,1
2,2
1,1
...
I wish to create a map of lists containing the y's for a given x... The result could look like this:
{1:[2,5,1], 2:[1,2]}
In python this would be straight forward to do in an imperative manner.. and would probably look somewhat like this:
d = defaultdict(list)
for x,y in csv_data:
d[x].append(y)
How would you go about achieving the same using functional programming techniques in F#?
Is it possible to do it as short, efficient and concise (and read-able) as in the given python example, using only functional style?, or would you have to fall back to imperative programming style with mutable data structures..?
Note: this is not a homework assignment, just me trying to wrap my head around functional programming
EDIT: My conclusion based on answers thus far
I tried timing each of the provided answers on a relative big csv file, just to get a feeling of the performance.. Furthermore I did a small test with the imperative approach:
let res = new Dictionary<string, List<string>>()
for row in l do
if (res.ContainsKey(fst row) = false) then
res.[fst row] <- new List<string>()
res.[fst row].Add(snd row)
The imperative approach completed in ~0.34 sec.
I think that the answer provided by Lee is the most general FP one, however the running time was ~4sec.
The answer given by Daniel ran in ~1.55sec.
And at last the answer given by jbtule ran in ~0.26. (I find it very interesting that it beat the imperative approach)
I used 'System.Diagnostics.Stopwatch()' for timing, and the code is executed as F# 3.0 in .Net 4.5
EDIT2: fixed stupid error in imperative f# code, and ensured that it uses the same list as the other solutions