learning-hmm 0.3.1.3 → 0.3.2.0

raw patch · 6 files changed

+56/−3 lines, 6 filesPVP ok

version bump matches the API change (PVP)

API changes (from Hackage documentation)

+ Learning.HMM: euclideanDistance :: (Eq s, Eq o) => HMM s o -> HMM s o -> Double

+ Learning.IOHMM: euclideanDistance :: (Eq i, Eq s, Eq o) => IOHMM i s o -> IOHMM i s o -> Double

Files

CHANGES.md +4/−0
learning-hmm.cabal +1/−1
src/Learning/HMM.hs +23/−0
src/Learning/HMM/Internal.hs +1/−1
src/Learning/IOHMM.hs +26/−0
src/Learning/IOHMM/Internal.hs +1/−1

CHANGES.md view

@@ -1,6 +1,10 @@ Revision history for Haskell package learning-hmm === +## Version 0.3.2.0+- Add function `euclideanDistance` which measures the Euclidean distance between+  two models+ ## Version 0.3.1.3 - Bug fix release

learning-hmm.cabal view

@@ -1,5 +1,5 @@ name:                learning-hmm-version:             0.3.1.3+version:             0.3.2.0 stability:           experimental  synopsis:            Yet another library for hidden Markov models

src/Learning/HMM.hs view

@@ -5,6 +5,7 @@   , LogLikelihood   , init   , withEmission+  , euclideanDistance   , viterbi   , baumWelch   , baumWelch'@@ -82,6 +83,16 @@     model'   = toInternal model     xs'      = U.fromList $ fromJust $ mapM (`V.elemIndex` outputs') xs +-- | Return the Euclidean distance between two models that have the same+--   states and outputs.+euclideanDistance :: (Eq s, Eq o) => HMM s o -> HMM s o -> Double+euclideanDistance model1 model2 =+  checkTwoModelsIn "euclideanDistance" model1 model2 `seq`+  I.euclideanDistance model1' model2'+  where+    model1' = toInternal model1+    model2' = toInternal model2+ -- | @viterbi model xs@ performs the Viterbi algorithm using the observed --   outputs @xs@, and returns the most likely state path and its log --   likelihood.@@ -144,6 +155,18 @@   | null states  = errorIn fun "empty states"   | null outputs = errorIn fun "empty outputs"   | otherwise    = ()++-- | Check if the two models have the same states and outputs.+checkTwoModelsIn :: (Eq s, Eq o) => String -> HMM s o -> HMM s o -> ()+checkTwoModelsIn fun model model'+  | ss /= ss' = errorIn fun "states disagree"+  | os /= os' = errorIn fun "outputs disagree"+  | otherwise = ()+  where+    ss  = states model+    ss' = states model'+    os  = outputs model+    os' = outputs model'  -- | Check if all the elements of the observed outputs are contained in the --   'outputs' of the model.

src/Learning/HMM/Internal.hs view

@@ -5,6 +5,7 @@   , LogLikelihood   , init   , withEmission+  , euclideanDistance   , viterbi   , baumWelch   , baumWelch'@@ -89,7 +90,6 @@      model' = fst $ head $ dropWhile ((> 1e-9) . snd) $ zip ms' ds --- | Return the Euclidean distance between two models. euclideanDistance :: HMM -> HMM -> Double euclideanDistance model model' =   sqrt $ H.sumElements ((w - w') ** 2) + H.sumElements ((phi - phi') ** 2)

src/Learning/IOHMM.hs view

@@ -5,6 +5,7 @@   , LogLikelihood   , init   , withEmission+  , euclideanDistance   , viterbi   , baumWelch   , baumWelch'@@ -93,6 +94,16 @@     xs'      = U.fromList $ fromJust $ mapM (`V.elemIndex` inputs') xs     ys'      = U.fromList $ fromJust $ mapM (`V.elemIndex` outputs') ys +-- | Return the Euclidean distance between two models that have the same+--   inputs, states, and outputs.+euclideanDistance :: (Eq i, Eq s, Eq o) => IOHMM i s o -> IOHMM i s o -> Double+euclideanDistance model1 model2 =+  checkTwoModelsIn "euclideanDistance" model1 model2 `seq`+  I.euclideanDistance model1' model2'+  where+    model1' = toInternal model1+    model2' = toInternal model2+ -- | @viterbi model xs ys@ performs the Viterbi algorithm using the inputs --   @xs@ and outputs @ys@, and returns the most likely state path and its --   log likelihood.@@ -167,6 +178,21 @@   | null states  = errorIn fun "empty states"   | null outputs = errorIn fun "empty outputs"   | otherwise    = ()++-- | Check if the two models have the same inputs, states, and outputs.+checkTwoModelsIn :: (Eq i, Eq s, Eq o) => String -> IOHMM i s o -> IOHMM i s o -> ()+checkTwoModelsIn fun model model'+  | is /= is' = errorIn fun "inputs disagree"+  | ss /= ss' = errorIn fun "states disagree"+  | os /= os' = errorIn fun "outputs disagree"+  | otherwise = ()+  where+    is  = inputs model+    is' = inputs model'+    ss  = states model+    ss' = states model'+    os  = outputs model+    os' = outputs model'  -- | Check if all the elements of the given inputs (outputs) are contained --   in the 'inputs' ('outputs') of the model.

src/Learning/IOHMM/Internal.hs view

@@ -5,6 +5,7 @@   , LogLikelihood   , init   , withEmission+  , euclideanDistance   , viterbi   , baumWelch   , baumWelch'@@ -93,7 +94,6 @@      model' = fst $ head $ dropWhile ((> 1e-9) . snd) $ zip ms' ds --- | Return the Euclidean distance between two models. euclideanDistance :: IOHMM -> IOHMM -> Double euclideanDistance model model' =   sqrt $ sum $ H.sumElements ((phi - phi') ** 2) :