Data.PrimitiveArray.ScoreMatrix

Simple score and distance matrices. These are two-dimensional tables together with row and column vectors of names.

Documentation

data ScoreMatrix t

NxN sized score matrices

TODO needs a vector with the column names!

Constructors

ScoreMatrix

Fields

• scoreMatrix :: !(Unboxed ((Z :. Int) :. Int) t)
• scoreNodes :: !(Unboxed Int t)
• rowNames :: !(Vector Text)
• colNames :: !(Vector Text)

Instances

(Unbox t, Eq t) => Eq (ScoreMatrix t)

Methods

(==) :: ScoreMatrix t -> ScoreMatrix t -> Bool

(/=) :: ScoreMatrix t -> ScoreMatrix t -> Bool

(Show t, Unbox t) => Show (ScoreMatrix t)

Methods

showsPrec :: Int -> ScoreMatrix t -> ShowS

show :: ScoreMatrix t -> String

showList :: [ScoreMatrix t] -> ShowS

(.!.) :: Unbox t => ScoreMatrix t -> (Int, Int) -> t

Get the distance between edges (From,To).

nodeDist :: Unbox t => ScoreMatrix t -> Int -> t

If the initial node has a "distance", it'll be here

rowNameOf :: ScoreMatrix t -> Int -> Text

Get the name of the node at an row index

colNameOf :: ScoreMatrix t -> Int -> Text

Get the name of the node at an column index

numRows :: Unbox t => ScoreMatrix t -> Int

Number of rows in a score matrix.

numCols :: Unbox t => ScoreMatrix t -> Int

Number of columns in a score matrix.

listOfRowNames :: ScoreMatrix t -> [Text]

listOfColNames :: ScoreMatrix t -> [Text]

toPartMatrix

Arguments

:: Double

temperature

-> ScoreMatrix Double

-> ScoreMatrix (Log Double)

Turns a ScoreMatrix for distances into one scaled by "temperature" for Inside/Outside calculations. Each value is scaled by k -> exp $ negate k / r * t where r = (n-1) * d d = mean of genetic distance

Node scores are turned directly into probabilities.

TODO Again, there is overlap and we should really have newtype Distance and friends.

TODO newtype Temperature = Temperature Double

TODO fix for rows /= cols!!!

fromFile :: FilePath -> IO (ScoreMatrix Double)

Import data.

TODO Should be generalized because Lib-BiobaseBlast does almost the same thing.