K-means algorithm in Haskell

KMeans.hs

-- This is the first thing I've ever written in Haskell by myself.
-- Be gentle.

import Data.List (genericLength)

type Assignment = [Int]

-- |Cluster data into k clusters
kmeans :: (Floating a, Ord a) => [[a]] -> Int -> Assignment
kmeans xs k = _kmeans xs $ nearestMeans xs (take k xs)

-- |Helper function for kmeans
_kmeans :: (Floating a, Ord a) => [[a]] -> Assignment -> Assignment
_kmeans xs asgn
    | newAsgn == asgn = asgn
    | otherwise = _kmeans xs newAsgn
    where newAsgn = nearestMeans xs $ calcMeans xs asgn

-- |Recompute the means
calcMeans :: (Floating a) => [[a]] -> Assignment -> [[a]]
calcMeans xs asgn = map listMean [[x | (x,i) <- zip xs asgn, i == j] | j <- [0..k]]
    where k = maximum asgn

-- |Mean of a list of lists
listMean :: (Floating a) => [[a]] -> [a]
listMean [] = error "mean of empty list "
listMean xs = map (/k) $ listSum xs
    where k = genericLength xs

-- |Sum of a list of lists
listSum :: (Floating a) => [[a]] -> [a]
listSum xs = foldl1 (\x1 x2 -> zipWith (+) x1 x2) xs

-- |Find the mean nearest each datum
nearestMeans :: (Floating a, Ord a) => [[a]] -> [[a]] -> [Int]
nearestMeans xs ms = map argmin dists
    where dists = [map (sumSquares x) ms | x <- xs]

-- |Calculate the sum of squared error between two lists
sumSquares :: (Floating a) => [a] -> [a] -> a
sumSquares xs ys = sum [(x-y)^2 | (x,y) <- zip xs ys]

-- |Get the index of the smallest element in a list
argmin :: (Ord a) => [a] -> Int
argmin (x:xs) = _argmin xs x 0 0

-- |Helper function for argmin
_argmin :: (Ord a) => [a] -> a -> Int -> Int -> Int
_argmin [x] val cix ix
  | x < val = cix + 1
  | otherwise = ix
_argmin (x:xs) val cix ix
  | x < val = _argmin xs x (cix+1) (cix+1)
  | otherwise = _argmin xs val (cix+1) ix

testData = [
       [ 19.44228191,  10.01954739],
       [ 19.53244135,   9.27926943],
       [ 19.47813524,  10.11702067],
       [ 19.53707403,  10.34821127],
       [ 20.22351335,  10.15144988],
       [ 19.28171056,   9.97469793],
       [ 20.02206671,  10.44740153],
       [ 20.50197584,  10.21343973],
       [ 19.21130219,  10.12152178],
       [ 20.12801058,   9.72609094],
       [ 19.83785357,  10.26330778],
       [ 20.41310726,   9.46206339],
       [ 19.45552404,   9.40547838],
       [ 19.89360546,  10.44014115],
       [ 20.36125385,   9.90567986],
       [ 20.35849112,  10.48796981],
       [ 20.88746346,  10.35747471],
       [ 20.49813333,  10.65385512],
       [ 20.62931967,  10.31353257],
       [ 19.71910217,  10.11721853],
       [ 23.88999494,   9.49780921],
       [ 23.53417881,  10.22200125],
       [ 24.20515374,  10.20352821],
       [ 24.27042841,   8.94589297],
       [ 24.33017732,  10.46796864],
       [ 24.1848164 ,   9.39154275],
       [ 24.25760608,  10.19827234],
       [ 24.14990874,   9.95071613],
       [ 23.79546847,   9.40557273],
       [ 24.12813273,   9.45732106],
       [ 24.88221126,  10.67626518],
       [ 24.24725526,   9.38441264],
       [ 23.70951565,   9.48528594],
       [ 23.84057452,  10.4090588 ],
       [ 25.30846271,  10.06039913],
       [ 23.88388447,   9.42833493],
       [ 23.77114657,   9.98375689],
       [ 23.40834507,  10.49376609],
       [ 23.98728878,  10.22910904],
       [ 22.96215117,   9.97424991],
       [ 19.58555678,  11.8102289 ],
       [ 20.03125014,  10.6541982 ],
       [ 20.61266943,  11.811877  ],
       [ 19.81763777,  12.87517492],
       [ 19.70835143,  11.64469843],
       [ 20.08416497,  11.49784966],
       [ 19.82591625,  11.04150988],
       [ 19.28604287,  12.57047196],
       [ 19.60033648,  12.37376337],
       [ 19.73600258,  11.13350926],
       [ 20.6532054 ,  12.40243406],
       [ 19.88155388,  11.5102145 ],
       [ 20.16608498,  11.52421444],
       [ 19.84189865,  12.16468441],
       [ 20.07146249,  11.9723343 ],
       [ 19.80540283,  12.25633198],
       [ 20.36592307,  12.04606745],
       [ 19.76154363,  12.04326451],
       [ 21.13411946,  11.97343609],
       [ 20.00825767,  12.22197451]]