implemented phase 2

DataModel.hs

@@ -11,8 +11,9 @@ class Freq a where
     frequency :: [ItemSet] -> a -> Frequency
 
 data Item = Item String deriving (Eq, Ord)
+
 instance Show Item where
-    show (Item s) = s --"Item " ++ s
+    show (Item a) = show a
 
 data ItemSet = ItemSet (Set Item) deriving (Eq, Ord)
 
@@ -27,16 +28,23 @@ instance Freq ItemSet where
 
 count :: [ItemSet] -> ItemSet -> Count
 count table (ItemSet set) =
-    length (filter isSuperSet table) where
+    length (filter isSuperset table) where
         isSuperset (ItemSet row) = set `Set.isSubsetOf` row
 
+difference :: ItemSet -> ItemSet -> ItemSet
+difference (ItemSet set1) (ItemSet set2) = ItemSet (Set.difference set1 set2)
+
+empty :: ItemSet
+empty = ItemSet (Set.fromList [])
+
 data Rule = Rule ItemSet ItemSet deriving (Eq)
 
 instance Show Rule where
     show (Rule a b) = show a ++ "-> " ++ show b
 
 instance Freq Rule where
-    frequency table (Rule (ItemSet set1) (ItemSet set2)) = frequency table $ ItemSet (set1 `Set.union` set2)
+    frequency table (Rule (ItemSet set1) (ItemSet set2)) = frequency table $
+        ItemSet (set1 `Set.union` set2)
 
 confidence :: [ItemSet] -> Rule -> Confidence
 confidence table (Rule x y) = frequency table (Rule x y) / frequency table x

ExtractRules.hs

@@ -0,0 +1,32 @@
+module ExtractRules (
+    extractRules
+) where
+import DataModel
+import qualified Data.Set as Set
+import qualified Data.Map as Map
+import Data.Map(Map)
+
+extractRules :: Confidence -> [ItemSet] -> [Rule]
+extractRules threshold patterns = filter (\x -> threshold <= confidence patterns x) rules where
+    rules = Map.foldrWithKey (\k v old -> ruleFromSubset v k : old) [] subsets
+    subsets = foldr (\x old -> insertMultiple (filteredPowerset x) x old) Map.empty patterns
+    filteredPowerset (ItemSet set) = map (ItemSet . Set.fromList) $
+        filter (\val -> val /= Set.toList set && val /= []) $ powerset $ Set.toList set
+
+ruleFromSubset :: ItemSet -> ItemSet -> Rule
+ruleFromSubset set subset = Rule subset (difference set subset)
+
+
+
+insertMultiple :: Ord k => [k] -> a -> Map k a -> Map k a
+insertMultiple keys value m = foldr (\x old -> Map.insert x value old) m keys
+
+powerset :: [a] -> [[a]]
+powerset []     = [[]]
+powerset (x:xs) = xss /\/ map (x:) xss where
+    xss = powerset xs
+
+(/\/) :: [a] -> [a] -> [a]
+[]     /\/ ys = ys
+(x:xs) /\/ ys = x : (ys /\/ xs)
+

FrequentPatterns.hs

@@ -7,9 +7,12 @@ import Debug.Trace (traceShow)
 import qualified Data.List as List
 
 semiUnion :: ItemSet -> ItemSet -> ItemSet
-semiUnion (ItemSet set1) (ItemSet set2) = ItemSet (if max1 <= max2 && Set.delete max1 set1 == Set.delete max2 set2 then set1 `Set.union` set2 else Set.empty) where
-    max1 = Set.findMax set1
-    max2 = Set.findMax set2
+semiUnion (ItemSet set1) (ItemSet set2) = ItemSet $
+    if max1 <= max2 && Set.delete max1 set1 == Set.delete max2 set2
+    then set1 `Set.union` set2
+    else Set.empty where
+        max1 = Set.findMax set1
+        max2 = Set.findMax set2
 
 -- generate all possible combinations from a set of singletons
 -- generateLevels :: [Item] -> [[ItemSet]]
@@ -19,15 +22,18 @@ semiUnion (ItemSet set1) (ItemSet set2) = ItemSet (if max1 <= max2 && Set.delete
 
 -- generate the next level in a bottom-up route
 generateNextLevel :: [ItemSet] -> [ItemSet]
-generateNextLevel level = traceShow ("Computing level " ++ show (isSize (head level))) $ foldr (\value old -> generate value ++ old) [] level where
-    generate value = takeWhile (/= empty) (foldr (\x old -> semiUnion value x : old) [] (tail $ List.dropWhile (/= value) level))
-    empty = ItemSet $ Set.fromList []
-    isSize (ItemSet set) = Set.size set
+generateNextLevel level = traceShow ("Computing level " ++ show (isSize (head level))) $
+    foldr (\value old -> generate value ++ old) [] level where
+        generate value = takeWhile (/= empty)
+            (foldr (\x old -> semiUnion value x : old) [] (tail $ List.dropWhile (/= value) level))
+        isSize (ItemSet set) = Set.size set
 
 singletons :: [ItemSet] -> [Item]
-singletons table = Set.toList $ foldr (\(ItemSet row) old -> old `Set.union` row) (Set.fromList []) table where
+singletons table = Set.toList $ foldr union (Set.fromList []) table where
+	union (ItemSet row) old = old `Set.union` row
 
 frequentPatterns :: Frequency -> [ItemSet] -> [[ItemSet]]
-frequentPatterns thresh table = until (\x -> [] == head x) (\x -> filterByFrequency (generateNextLevel (head x)) : x) [firstLevel] where
-    firstLevel = filterByFrequency $ map (\x -> ItemSet $ Set.fromList [x]) (singletons table)
-    filterByFrequency = filter (\x -> frequency table x >= thresh)
+frequentPatterns thresh table = until (\x -> [] == head x)
+    (\x -> filterByFrequency (generateNextLevel (head x)) : x) [firstLevel] where
+        firstLevel = filterByFrequency $ map (\x -> ItemSet $ Set.fromList [x]) (singletons table)
+        filterByFrequency = filter (\x -> frequency table x >= thresh)

phase1.hs

@@ -13,6 +13,6 @@ main = do
     let threshold = read $ last args
     file <- readFile filename
     case parseCSV file of
-    Left _ -> error "Could not parse out.csv"
-    Right val -> mapM_ (\x -> putStrLn (show x ++ "(" ++ show (count table x) ++ ")")) (concat (frequentPatterns threshold table)) where
-        table = map (ItemSet. Set.fromList .map Item) val
+        Left _ -> error "Could not parse out.csv"
+        Right val -> mapM_ (\x -> putStrLn (show x ++ "(" ++ show (count table x) ++ ")")) (concat (frequentPatterns threshold table)) where
+            table = map (ItemSet. Set.fromList .map Item) val

phase2.hs

@@ -0,0 +1,20 @@
+import CSVParser
+import FrequentPatterns
+import DataModel
+import ExtractRules
+import qualified Data.Set as Set
+import System.Environment(getArgs)
+import Control.Monad(when)
+
+main :: IO()
+main = do
+    args <- getArgs
+    when (2 /= length args) (error "Usage: phase2 <file.csv> <threshold>")
+    let filename = head args
+    let threshold = read $ last args
+    file <- readFile filename
+    case parseCSV file of
+        Left _ -> error "Could not parse out.csv"
+        Right val -> print $ extractRules threshold $
+                filter (/= empty) $ concat $ frequentPatterns 0.01 table where
+                    table = map (ItemSet. Set.fromList .map Item) val