Safe Haskell | None |
---|---|
Language | Haskell2010 |
Synopsis
- constant :: y -> T y
- linear :: C y => y -> y -> T y
- linearMultiscale :: C y => y -> y -> T y
- linearMultiscaleNeutral :: C y => y -> T y
- linearStable :: C y => y -> y -> T y
- linearMean :: C y => y -> y -> T y
- line :: C y => Int -> (y, y) -> T y
- exponential :: C y => y -> y -> T y
- exponentialMultiscale :: C y => y -> y -> T y
- exponentialStable :: C y => y -> y -> T y
- exponentialMultiscaleNeutral :: C y => y -> T y
- exponential2 :: C y => y -> y -> T y
- exponential2Multiscale :: C y => y -> y -> T y
- exponential2Stable :: C y => y -> y -> T y
- exponential2MultiscaleNeutral :: C y => y -> T y
- exponentialFromTo :: C y => y -> y -> y -> T y
- exponentialFromToMultiscale :: C y => y -> y -> y -> T y
- vectorExponential :: (C y, C y v) => y -> v -> T v
- vectorExponential2 :: (C y, C y v) => y -> v -> T v
- cosine :: C y => y -> y -> T y
- cosineMultiscale :: C y => y -> y -> T y
- cosineSubdiv :: C y => y -> y -> T y
- cosineStable :: C y => y -> y -> T y
- cubicHermite :: C y => (y, (y, y)) -> (y, (y, y)) -> T y
- cubicHermiteStable :: C y => (y, (y, y)) -> (y, (y, y)) -> T y
- curveMultiscale :: (y -> y -> y) -> y -> y -> T y
- curveMultiscaleNeutral :: (y -> y -> y) -> y -> y -> T y
- cubicFunc :: C y => (y, (y, y)) -> (y, (y, y)) -> y -> y
- cosineWithSlope :: C y => (y -> y -> signal) -> y -> y -> signal
Documentation
linearMultiscale :: C y => y -> y -> T y Source #
Minimize rounding errors by reducing number of operations per element to a logarithmuc number.
linearMultiscaleNeutral :: C y => y -> T y Source #
Linear curve starting at zero.
linearStable :: C y => y -> y -> T y Source #
As stable as the addition of time values.
linearMean :: C y => y -> y -> T y Source #
It computes the same like linear
but in a numerically more stable manner,
namely using a subdivision scheme.
The division needed is a division by two.
0 4 8 0 2 4 6 8 0 1 2 3 4 5 6 7 8
Linear curve of a fixed length. The final value is not actually reached, instead we stop one step before. This way we can concatenate several lines without duplicate adjacent values.
:: (C y, C y v) | |
=> y | time where the function reaches 1/e of the initial value |
-> v | initial value |
-> T v | exponential decay |
This is an extension of exponential
to vectors
which is straight-forward but requires more explicit signatures.
But since it is needed rarely I setup a separate function.
cubicHermite :: C y => (y, (y, y)) -> (y, (y, y)) -> T y Source #
cubicHermiteStable :: C y => (y, (y, y)) -> (y, (y, y)) -> T y Source #
curveMultiscale :: (y -> y -> y) -> y -> y -> T y Source #
curveMultiscaleNeutral :: (y -> y -> y) -> y -> y -> T y Source #
cubicFunc :: C y => (y, (y, y)) -> (y, (y, y)) -> y -> y Source #
0 16 0 8 16 0 4 8 12 16 0 2 4 6 8 10 12 14 16 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
cosineWithSlope :: C y => (y -> y -> signal) -> y -> y -> signal Source #