diff --git a/ChangeLog.md b/ChangeLog.md --- a/ChangeLog.md +++ b/ChangeLog.md @@ -5,6 +5,13 @@ ## Unreleased changes +## Version 0.6.1.0 + +- Split `ELynx.Tools` into separate modules because the package will be reduced. +- Remove the following modules from `ELynx.Tools`: `Concurrent`, + `LinearAlgebra`, `List`, `Misc`, and `Numeric`. + + ## Version 0.6.0.0 - **elynx-tree:** remove parallel folds with layers (`parBranchFoldMapWithLayer` diff --git a/README.md b/README.md --- a/README.md +++ b/README.md @@ -69,13 +69,13 @@ # Get help -For example: - - slynx --help + cabal exec slynx -- --help + # OR: stack exec slynx -- --help + # OR: slynx --help ELynx Suite version 0.6.0.0. Developed by Dominik Schrempf. - Compiled on September 3, 2021, at 20:56 pm, UTC. + Compiled on September 4, 2021, at 12:58 pm, UTC. Usage: slynx [-v|--verbosity VALUE] [-o|--output-file-basename NAME] [-f|--force] [--no-elynx-file] COMMAND @@ -113,6 +113,7 @@ - ProteinX (amino acids; including gaps) - ProteinS (amino acids; including gaps, and translation stops) - ProteinI (amino acids; including gaps, translation stops, and IUPAC codes) + ELynx ----- A Haskell library and tool set for computational biology. The goal of ELynx is @@ -126,6 +127,9 @@ tlynx Analyze, modify, and simulate phylogenetic trees. elynx Validate and redo past analyses. + Get help for commands: + slynx --help + Get help for sub commands: slynx examine --help @@ -134,11 +138,13 @@ The documentation of sub commands can be accessed separately: - slynx simulate --help + cabal exec slynx -- simulate --help + # OR: stack exec slynx -- simulate --help + # OR: slynx simulate --help ELynx Suite version 0.6.0.0. Developed by Dominik Schrempf. - Compiled on September 3, 2021, at 20:56 pm, UTC. + Compiled on September 4, 2021, at 12:58 pm, UTC. Usage: slynx simulate (-t|--tree-file Name) [-s|--substitution-model MODEL] [-m|--mixture-model MODEL] [-e|--edm-file NAME] diff --git a/src/TLynx/Compare/Compare.hs b/src/TLynx/Compare/Compare.hs --- a/src/TLynx/Compare/Compare.hs +++ b/src/TLynx/Compare/Compare.hs @@ -26,7 +26,10 @@ import qualified Data.Set as S import qualified Data.Text as T import qualified Data.Text.IO as T -import ELynx.Tools +import ELynx.Tools.ELynx +import ELynx.Tools.Environment +import ELynx.Tools.Logger +import ELynx.Tools.Options import ELynx.Tree import Graphics.Gnuplot.Simple import System.IO diff --git a/src/TLynx/Compare/Options.hs b/src/TLynx/Compare/Options.hs --- a/src/TLynx/Compare/Options.hs +++ b/src/TLynx/Compare/Options.hs @@ -17,7 +17,9 @@ ) where -import ELynx.Tools +import Data.Aeson +import ELynx.Tools.Reproduction +import GHC.Generics import Options.Applicative import TLynx.Parsers diff --git a/src/TLynx/Connect/Connect.hs b/src/TLynx/Connect/Connect.hs --- a/src/TLynx/Connect/Connect.hs +++ b/src/TLynx/Connect/Connect.hs @@ -22,7 +22,9 @@ import qualified Data.ByteString.Lazy.Char8 as BL import Data.Default.Class import qualified Data.Set as S -import ELynx.Tools +import ELynx.Tools.ELynx +import ELynx.Tools.Environment +import ELynx.Tools.Logger import ELynx.Tree import System.IO import TLynx.Connect.Options diff --git a/src/TLynx/Connect/Options.hs b/src/TLynx/Connect/Options.hs --- a/src/TLynx/Connect/Options.hs +++ b/src/TLynx/Connect/Options.hs @@ -17,7 +17,9 @@ ) where -import ELynx.Tools +import Data.Aeson +import ELynx.Tools.Reproduction +import GHC.Generics import Options.Applicative import TLynx.Parsers diff --git a/src/TLynx/Distance/Distance.hs b/src/TLynx/Distance/Distance.hs --- a/src/TLynx/Distance/Distance.hs +++ b/src/TLynx/Distance/Distance.hs @@ -33,7 +33,10 @@ import qualified Data.Text as T import qualified Data.Text.IO as T import qualified Data.Vector.Unboxed as V -import ELynx.Tools +import ELynx.Tools.ByteString +import ELynx.Tools.ELynx +import ELynx.Tools.Environment +import ELynx.Tools.Logger import ELynx.Tree import Statistics.Sample import System.IO diff --git a/src/TLynx/Distance/Options.hs b/src/TLynx/Distance/Options.hs --- a/src/TLynx/Distance/Options.hs +++ b/src/TLynx/Distance/Options.hs @@ -20,10 +20,12 @@ ) where +import Data.Aeson import qualified Data.Attoparsec.ByteString.Char8 as AC import qualified Data.ByteString.Char8 as BS -import ELynx.Tools +import ELynx.Tools.Reproduction import ELynx.Tree (Support (..), toSupportUnsafe) +import GHC.Generics import Options.Applicative import TLynx.Parsers import Text.Printf diff --git a/src/TLynx/Examine/Examine.hs b/src/TLynx/Examine/Examine.hs --- a/src/TLynx/Examine/Examine.hs +++ b/src/TLynx/Examine/Examine.hs @@ -22,7 +22,10 @@ import qualified Data.ByteString.Lazy.Char8 as BL import Data.Containers.ListUtils (nubOrd) import Data.List ((\\)) -import ELynx.Tools +import ELynx.Tools.ByteString +import ELynx.Tools.ELynx +import ELynx.Tools.Environment +import ELynx.Tools.Logger import ELynx.Tree import System.IO ( Handle, diff --git a/src/TLynx/Examine/Options.hs b/src/TLynx/Examine/Options.hs --- a/src/TLynx/Examine/Options.hs +++ b/src/TLynx/Examine/Options.hs @@ -17,7 +17,9 @@ ) where -import ELynx.Tools +import Data.Aeson +import ELynx.Tools.Reproduction +import GHC.Generics import Options.Applicative import TLynx.Parsers diff --git a/src/TLynx/Grabble.hs b/src/TLynx/Grabble.hs new file mode 100644 --- /dev/null +++ b/src/TLynx/Grabble.hs @@ -0,0 +1,41 @@ +-- | +-- Module : TLynx.Grabble +-- Description : Grabble a list +-- Copyright : (c) 2021 Dominik Schrempf +-- License : GPL-3.0-or-later +-- +-- Maintainer : dominik.schrempf@gmail.com +-- Stability : experimental +-- Portability : portable +-- +-- Creation date: Mon Sep 6 09:39:16 2021. +module TLynx.Grabble + ( grabble, + ) +where + +import Control.Monad +import Control.Monad.Primitive +import Control.Monad.ST +import Data.Vector (Vector) +import qualified Data.Vector as V +import qualified Data.Vector.Mutable as M +import System.Random.MWC + +-- | @grabble xs m n@ is /O(m*n')/, where @n' = min n (length xs)@. Choose @n'@ +-- elements from @xs@, without replacement, and that @m@ times. +grabble :: PrimMonad m => [a] -> Int -> Int -> Gen (PrimState m) -> m [[a]] +grabble xs m n gen = do + swapss <- replicateM m $ + forM [0 .. min (l - 1) n] $ \i -> do + j <- uniformR (i, l) gen + return (i, j) + return $ map (V.toList . V.take n . swapElems (V.fromList xs)) swapss + where + l = length xs - 1 + +swapElems :: Vector a -> [(Int, Int)] -> Vector a +swapElems xs swaps = runST $ do + mxs <- V.unsafeThaw xs + mapM_ (uncurry $ M.unsafeSwap mxs) swaps + V.unsafeFreeze mxs diff --git a/src/TLynx/Options.hs b/src/TLynx/Options.hs --- a/src/TLynx/Options.hs +++ b/src/TLynx/Options.hs @@ -11,14 +11,12 @@ -- Portability : portable -- -- Creation date: Sat Sep 7 18:55:03 2019. -module TLynx.Options - ( Arguments (..), - CommandArguments (..), - parseArguments, - ) -where +module TLynx.Options (CommandArguments (..)) where -import ELynx.Tools +import Data.Aeson +import ELynx.Tools.Options +import ELynx.Tools.Reproduction +import GHC.Generics import Options.Applicative import TLynx.Compare.Options import TLynx.Connect.Options diff --git a/src/TLynx/Parsers.hs b/src/TLynx/Parsers.hs --- a/src/TLynx/Parsers.hs +++ b/src/TLynx/Parsers.hs @@ -19,7 +19,7 @@ where import Data.List -import ELynx.Tools +import ELynx.Tools.InputOutput import ELynx.Tree import Options.Applicative @@ -74,7 +74,7 @@ ++ "; default: Standard; for detailed help, see 'tlynx --help'" ) where - nwfs = map show (allValues :: [NewickFormat]) + nwfs = map show ([minBound ..] :: [NewickFormat]) nwlist = intercalate ", " (init nwfs) <> ", or " <> last nwfs -- | Help for different 'NewickFormat's. @@ -82,7 +82,7 @@ newickHelp = map (toListItem . describeNewickFormat) - (allValues :: [NewickFormat]) + ([minBound ..] :: [NewickFormat]) ++ ["- Nexus file including Newick trees"] where toListItem = ("- Newick " ++) diff --git a/src/TLynx/Shuffle/Options.hs b/src/TLynx/Shuffle/Options.hs --- a/src/TLynx/Shuffle/Options.hs +++ b/src/TLynx/Shuffle/Options.hs @@ -17,7 +17,10 @@ ) where -import ELynx.Tools +import Data.Aeson +import ELynx.Tools.Options +import ELynx.Tools.Reproduction +import GHC.Generics import Options.Applicative import TLynx.Parsers diff --git a/src/TLynx/Shuffle/Shuffle.hs b/src/TLynx/Shuffle/Shuffle.hs --- a/src/TLynx/Shuffle/Shuffle.hs +++ b/src/TLynx/Shuffle/Shuffle.hs @@ -25,7 +25,11 @@ import Control.Monad.IO.Class (liftIO) import Control.Monad.Trans.Reader (ask) import qualified Data.ByteString.Lazy.Char8 as BL -import ELynx.Tools +import ELynx.Tools.Definitions +import ELynx.Tools.ELynx +import ELynx.Tools.Environment +import ELynx.Tools.Logger +import ELynx.Tools.Reproduction import ELynx.Tree import ELynx.Tree.Simulate.PointProcess ( PointProcess (PointProcess), @@ -33,6 +37,7 @@ ) import System.IO (hClose) import System.Random.MWC (GenIO, initialize) +import TLynx.Grabble import TLynx.Parsers import TLynx.Shuffle.Options diff --git a/src/TLynx/Simulate/Options.hs b/src/TLynx/Simulate/Options.hs --- a/src/TLynx/Simulate/Options.hs +++ b/src/TLynx/Simulate/Options.hs @@ -22,10 +22,13 @@ ) where +import Data.Aeson import Data.List import Data.Maybe -import ELynx.Tools +import ELynx.Tools.Options +import ELynx.Tools.Reproduction import ELynx.Tree.Simulate.PointProcess (TimeSpec (..)) +import GHC.Generics import Options.Applicative deriving instance Eq TimeSpec diff --git a/src/TLynx/Simulate/Simulate.hs b/src/TLynx/Simulate/Simulate.hs --- a/src/TLynx/Simulate/Simulate.hs +++ b/src/TLynx/Simulate/Simulate.hs @@ -1,4 +1,5 @@ {-# LANGUAGE OverloadedStrings #-} +{-# LANGUAGE ScopedTypeVariables #-} {-# LANGUAGE TemplateHaskell #-} {-# LANGUAGE TupleSections #-} @@ -31,6 +32,7 @@ ) import Control.Monad import Control.Monad.IO.Class +import Control.Monad.Primitive import Control.Monad.Trans.Reader hiding (local) import Control.Parallel.Strategies import qualified Data.ByteString.Builder as BB @@ -39,12 +41,36 @@ import Data.Maybe import qualified Data.Sequence as Seq import qualified Data.Set as Set -import ELynx.Tools +import qualified Data.Vector.Unboxed as VU +import Data.Word +import ELynx.Tools.ELynx +import ELynx.Tools.Environment +import ELynx.Tools.Logger +import ELynx.Tools.Reproduction import ELynx.Tree import qualified ELynx.Tree.Simulate.Coalescent as CS import qualified ELynx.Tree.Simulate.PointProcess as PP import System.Random.MWC +import TLynx.Grabble import TLynx.Simulate.Options + +-- Split a generator. +splitGen :: PrimMonad m => Int -> Gen (PrimState m) -> m [Gen (PrimState m)] +splitGen n gen + | n <= 0 = return [] + | otherwise = do + seeds :: [VU.Vector Word32] <- replicateM (n -1) $ uniformVector gen 256 + fmap (gen :) (mapM initialize seeds) + +-- For a given number of capabilities and number of calculations, get chunk +-- sizes. The chunk sizes will be as evenly distributed as possible and sum up +-- to the number of calculations. +getChunks :: Int -> Int -> [Int] +getChunks c n = ns + where + n' = n `div` c + r = n `mod` c + ns = replicate r (n' + 1) ++ replicate (c - r) n' -- | Simulate phylogenetic trees using birth and death process. simulate :: ELynx SimulateArguments () diff --git a/src/TLynx/TLynx.hs b/src/TLynx/TLynx.hs --- a/src/TLynx/TLynx.hs +++ b/src/TLynx/TLynx.hs @@ -15,7 +15,8 @@ ) where -import ELynx.Tools +import ELynx.Tools.ELynx +import ELynx.Tools.Options import TLynx.Compare.Compare import TLynx.Connect.Connect import TLynx.Distance.Distance diff --git a/tlynx.cabal b/tlynx.cabal --- a/tlynx.cabal +++ b/tlynx.cabal @@ -1,6 +1,6 @@ cabal-version: 2.2 name: tlynx -version: 0.6.0.0 +version: 0.6.1.0 synopsis: Handle phylogenetic trees description: Examine, compare, and simulate phylogenetic trees in a reproducible way. Please see the README on GitHub at . category: Bioinformatics @@ -31,6 +31,7 @@ TLynx.Distance.Options TLynx.Examine.Examine TLynx.Examine.Options + TLynx.Grabble TLynx.Options TLynx.Parsers TLynx.Shuffle.Options @@ -59,6 +60,7 @@ , mwc-random , optparse-applicative , parallel + , primitive , statistics , text , transformers