ekmett/ad

Issues

• Fix test suite on `x86_64-darwin`

  #113 opened 6 days ago by RyanGlScott
  1

• Tape is only exported by Internal modules

  #112 opened a month ago by jmcarthur
  0

• Difference between `auto` and `realToFrac`

  #98 opened 3 months ago by dschrempf
  1

• Incorrect Results for Extra Floating Functions

  #108 opened 3 months ago by julmb
  19

• Discrepancies Between Different Modes

  #109 opened 3 months ago by julmb
  1

• Avoid uses of `head` and `tail`

  #107 opened 9 months ago by RyanGlScott
  0

• Implement `fast-math` Version of `ReverseDouble` with `+ffi`

  #106 opened 10 months ago by julmb
  4

• `ReverseDouble` with `+ffi` produces different result for square root of product

  #104 opened 10 months ago by julmb
  17

• Performance of `Numeric.AD.Mode.Reverse.Double`

  #96 opened 3 years ago by julmb
  2

• Not building with GHC 9.6.1-alpha3

  #102 opened a year ago by konn
  1

• Build failure with GHC 9.6.1-alpha1

  #100 opened a year ago by RyanGlScott
  0

• Discrepancy between `Reverse` and `ReverseDouble` when using flag `+ffi`

  #97 opened 2 years ago by julmb
  4

• Matrix algebra support?

  #70 opened 6 years ago by jluttine
  4

• How can I add custom gradients?

  #77 opened 5 years ago by kai-qu
  11

• nesting: diff (\x -> diff (*x) 1) 1

  #60 opened 8 years ago by barak
  5

• du failing on simple function

  #38 opened 10 years ago by barak
  13

• Two associated pattern synonyms (KnownConstant / KnownZero) for Mode

  #65 opened 3 years ago by Icelandjack
  1

• zero, NaN, zero?

  #71 opened 3 years ago by barak
  0

• Using AD with exceptions

  #69 opened 3 years ago by mbernat
  1

• docspec fails with GHC-7.4.2

  #92 opened 3 years ago by phadej
  2

• doctest failure with data-reify-0.6.2 (but not with data-reify-0.6.1)

  #89 opened 4 years ago by RyanGlScott
  2

• Feasibility finder often loops on Newton constrainedDescent

  #78 opened 5 years ago by identicalsnowflake
  0

• Test suite floating point accuracy failure when building with non-glibc-libc

  #73 opened 5 years ago by nh2
  0

• Jacobian fails when function is a parameter

  #74 opened 5 years ago by Magalame
  10

• Nesting Benchmark

  #72 opened 6 years ago by barak
  1

• Could not deduce (Semigroup (Id a))

  #68 opened 6 years ago by phadej
  1

• Unexpected difference in behavior of `grad` for different modes?

  #54 opened 7 years ago by mstksg
  13

• weird stuff with partial function application to lists?

  #66 opened 7 years ago by donovancrichton
  1

• Multi-variate root finder?

  #59 opened 8 years ago by augustss
  2

• AD with Accelerate types

  #63 opened 7 years ago by alang9
  0

• doctest errors with GHC-8.1.20170123

  #62 opened 7 years ago by phadej
  4

• GHC 8 compatible release needed

  #58 opened 8 years ago by alexanderkjeldaas
  5

• Release 4.3.2 on Hackage

  #53 opened 8 years ago by expipiplus1
  1

• Gradient descent fails to converge (to the right result).

  #45 opened 9 years ago by echatav
  7

• No documentation on hackage for 4.3.1

  #49 opened 9 years ago by expipiplus1
  1

• Use `diffs` instead of `diffs0` in `findZero`?

  #50 opened 9 years ago by ekmett
  1

• Double specific implementation of Newton?

  #46 opened 9 years ago by ocharles
  3

• miscalculating grads

  #41 opened 9 years ago by echatav
  5

• 'cabal install AD' fails when installing semigroupoids

  #44 opened 9 years ago by fouronnes
  4

• Determining when Jacobians are sparse

  #40 opened 9 years ago by dmcclean
  21

• fixup opening docs for Numeric.AD.Internal.Reverse

  #37 opened 9 years ago by cartazio
  0

• Add stochastic gradient descent

  #24 opened 10 years ago by ekmett
  5

• 4.2 compilation failure

  #36 opened 10 years ago by snoyberg
  1

• Horribly broken type signatures in Numeric.AD.Newton in 4

  #35 opened 10 years ago by ekmett
  4

• conjugateGradient used to give the wrong answer, now it gives no answer...

  #28 opened 10 years ago by barak
  11

• Readme refers to "tensors" but it is no more

  #29 opened 10 years ago by jfischoff
  1

• Fix ComposeMode Data/Typeable

  #33 opened 10 years ago by ekmett
  0

• Missing instances.h

  #34 opened 10 years ago by snoyberg
  1

• Bug in conjugateGradientDescent

  #26 opened 11 years ago by cscherrer
  1

• Generate Specialized Code for known objective functions (recording prior dev discussion on this topic)

  #25 opened 11 years ago by cartazio
  0