]>
gitweb.michael.orlitzky.com - numerical-analysis.git/log
Michael Orlitzky [Fri, 7 Dec 2018 14:37:41 +0000 (09:37 -0500)]
numerical-analysis.cabal: add BigFloat to exposed-modules.
Michael Orlitzky [Fri, 7 Dec 2018 14:32:22 +0000 (09:32 -0500)]
numerical-analysis.cabal: update fixed-vector and doctest dependencies.
Michael Orlitzky [Tue, 15 Apr 2014 14:50:14 +0000 (10:50 -0400)]
Fix a bug in the relative error.
Michael Orlitzky [Mon, 14 Apr 2014 03:20:11 +0000 (23:20 -0400)]
Remove some commented code from FEM.R1.
Michael Orlitzky [Mon, 14 Apr 2014 03:17:35 +0000 (23:17 -0400)]
Add relative error functions to FEM.R1.
Michael Orlitzky [Mon, 14 Apr 2014 03:17:05 +0000 (23:17 -0400)]
Add natural numbers up to N64.
Michael Orlitzky [Sun, 13 Apr 2014 19:55:49 +0000 (15:55 -0400)]
Speed up the Cholesky factorization with a fold.
Michael Orlitzky [Sun, 13 Apr 2014 18:12:59 +0000 (14:12 -0400)]
Add column synonyms up to Col32.
Michael Orlitzky [Sun, 13 Apr 2014 18:09:17 +0000 (14:09 -0400)]
Add natural numbers up to N32.
Michael Orlitzky [Sun, 13 Apr 2014 08:12:52 +0000 (04:12 -0400)]
Fix bugs in big_K, big_M.
Michael Orlitzky [Mon, 7 Apr 2014 16:57:37 +0000 (12:57 -0400)]
Add Mat6 and Mat7 synonyms.
Move the example FEM into its own FEM.R1.Exampl1 module.
Use the new example FEM module in the FEM.R1 doctests.
Michael Orlitzky [Mon, 7 Apr 2014 16:16:40 +0000 (12:16 -0400)]
Remove solution', because what did it do?
Michael Orlitzky [Mon, 7 Apr 2014 16:09:23 +0000 (12:09 -0400)]
Fix a few warnings.
Michael Orlitzky [Mon, 7 Apr 2014 15:47:14 +0000 (11:47 -0400)]
Commit whatever I was working on last and have forgotten about.
Michael Orlitzky [Wed, 19 Feb 2014 20:36:28 +0000 (15:36 -0500)]
Add a FEM.R1 example (should be moved into the tree after being cleaned up).
Michael Orlitzky [Wed, 19 Feb 2014 20:35:41 +0000 (15:35 -0500)]
Finish the FEM.R1 solution for Dirichlet boundary conditions.
Michael Orlitzky [Wed, 19 Feb 2014 01:26:13 +0000 (20:26 -0500)]
Require a matrix to be nonempty if we're going to QR factorize it.
Michael Orlitzky [Wed, 19 Feb 2014 01:25:05 +0000 (20:25 -0500)]
Generalize the signature of zipwith2.
Implement the dot product using zipwith2.
Implement matrix multiplication using the dot product.
Implement the frobenius_norm with zipwith2.
Michael Orlitzky [Wed, 19 Feb 2014 01:22:12 +0000 (20:22 -0500)]
Use the element_sum2 function to sum the terms in Gaussian integration instead of the incorrect 1-norm.
Michael Orlitzky [Wed, 19 Feb 2014 00:53:35 +0000 (19:53 -0500)]
Fix the type of zip2 to be more general.
Michael Orlitzky [Wed, 19 Feb 2014 00:52:44 +0000 (19:52 -0500)]
Use the Col type synonym in the column Normed instance.
Michael Orlitzky [Wed, 19 Feb 2014 00:51:18 +0000 (19:51 -0500)]
Fix the column matrix 1-norm (it was missing the absolute value).
Michael Orlitzky [Wed, 19 Feb 2014 00:50:32 +0000 (19:50 -0500)]
Add the list_plot' function to Piecewise.
Michael Orlitzky [Sun, 16 Feb 2014 02:27:34 +0000 (21:27 -0500)]
Add 'coefficients' to FEM.R1 which solves the desired system.
Michael Orlitzky [Sun, 16 Feb 2014 02:27:14 +0000 (21:27 -0500)]
Add a preliminary Piecewise module.
Michael Orlitzky [Sun, 16 Feb 2014 02:26:46 +0000 (21:26 -0500)]
Define Col6...Col9 types in Linear.Matrix.
Michael Orlitzky [Sun, 16 Feb 2014 02:26:19 +0000 (21:26 -0500)]
Export everything from Z to N20 from Naturals.
Michael Orlitzky [Wed, 12 Feb 2014 06:52:35 +0000 (01:52 -0500)]
Fix the incorrect definition of "minor".
Add a "preminor" function that does what "minor" used to.
Implement cofactors and a matrix inverse in terms of them.
Michael Orlitzky [Wed, 12 Feb 2014 05:00:27 +0000 (00:00 -0500)]
Add big_K code to FEM.R1.
Michael Orlitzky [Wed, 12 Feb 2014 01:14:47 +0000 (20:14 -0500)]
Use a better implementation of backwards_substitute.
Michael Orlitzky [Wed, 12 Feb 2014 00:21:53 +0000 (19:21 -0500)]
Add Linear.Matrix.ifoldr2.
Michael Orlitzky [Tue, 11 Feb 2014 21:10:53 +0000 (16:10 -0500)]
Implement forward substitute in terms of a fold.
Michael Orlitzky [Tue, 11 Feb 2014 21:10:18 +0000 (16:10 -0500)]
Add Linear.Matrix.set_idx.
Loosen constraints on the dot product function.
Michael Orlitzky [Tue, 11 Feb 2014 20:25:49 +0000 (15:25 -0500)]
Update Integration.Gaussian to use zip2.
Michael Orlitzky [Tue, 11 Feb 2014 20:22:35 +0000 (15:22 -0500)]
Generalize colzip and colzipwith to any matrices.
Add zip2three to zip three matrices together.
Michael Orlitzky [Tue, 11 Feb 2014 20:07:53 +0000 (15:07 -0500)]
Fix backward_substitute.
Add 'solve_positive_definite' to Linear.System.
Add tests for everything in Linear.System.
Michael Orlitzky [Tue, 11 Feb 2014 20:07:35 +0000 (15:07 -0500)]
Add reverse2 to Linear.Matrix.
Michael Orlitzky [Tue, 11 Feb 2014 18:16:26 +0000 (13:16 -0500)]
Add Eq constraints throughout Linear.Iteration.
Michael Orlitzky [Tue, 11 Feb 2014 18:15:15 +0000 (13:15 -0500)]
Add another test for 'cholesky'.
Improve implementation of upper/lower triangular tests.
Drop some typeclass constraints on upper/lower triangular tests.
Michael Orlitzky [Sat, 8 Feb 2014 00:33:45 +0000 (19:33 -0500)]
Add the imap2 function to Linear.Matrix.
Michael Orlitzky [Sat, 8 Feb 2014 00:24:14 +0000 (19:24 -0500)]
Make safe indexing slower but not dependent on the number of rows/columns.
Michael Orlitzky [Sat, 8 Feb 2014 00:10:02 +0000 (19:10 -0500)]
Remove the "row" function which returned a vector instead of a matrix.
Michael Orlitzky [Fri, 7 Feb 2014 23:56:21 +0000 (18:56 -0500)]
Drop the 'column' function that returned a vector instead of a matrix.
Michael Orlitzky [Fri, 7 Feb 2014 23:46:51 +0000 (18:46 -0500)]
Rename matmap to map2.
Michael Orlitzky [Fri, 7 Feb 2014 15:20:26 +0000 (10:20 -0500)]
More progress on the 1d FEM, the "big F" vector is now computed correctly.
Michael Orlitzky [Fri, 7 Feb 2014 15:19:53 +0000 (10:19 -0500)]
Add the 'ifoldl2' function to Linear.Matrix.
Michael Orlitzky [Fri, 7 Feb 2014 15:15:39 +0000 (10:15 -0500)]
Fix a typo.
Michael Orlitzky [Fri, 7 Feb 2014 15:15:11 +0000 (10:15 -0500)]
Add a bunch more natural numbers in the Naturals module.
Michael Orlitzky [Thu, 6 Feb 2014 06:36:45 +0000 (01:36 -0500)]
Add the (in progress) FEM.R1 module.
Michael Orlitzky [Thu, 6 Feb 2014 06:36:28 +0000 (01:36 -0500)]
Fix some docs in Roots.Simple.
Michael Orlitzky [Wed, 5 Feb 2014 20:30:47 +0000 (15:30 -0500)]
Implement Gaussian quadrature and test it.
Michael Orlitzky [Wed, 5 Feb 2014 20:30:26 +0000 (15:30 -0500)]
Rename zipcol to colzip.
Add a new function, colzipwith.
Michael Orlitzky [Wed, 5 Feb 2014 01:33:31 +0000 (20:33 -0500)]
Add the Integration.Gaussian.nodes_and_weights function.
Michael Orlitzky [Wed, 5 Feb 2014 01:33:09 +0000 (20:33 -0500)]
Add zipcol and matmap functions in Linear.Matrix.
Michael Orlitzky [Tue, 4 Feb 2014 22:41:19 +0000 (17:41 -0500)]
Add the Integration.Gaussian module where we begin to implement Gaussian quadrature.
Michael Orlitzky [Tue, 4 Feb 2014 22:40:59 +0000 (17:40 -0500)]
Add comments to Linear.QR about non-convergence.
Michael Orlitzky [Tue, 4 Feb 2014 20:03:56 +0000 (15:03 -0500)]
Finish the eigenvectors_symmetric tests.
Michael Orlitzky [Tue, 4 Feb 2014 17:16:52 +0000 (12:16 -0500)]
Add the eigenvectors_symmetric function and begin writing tests for it.
Michael Orlitzky [Tue, 4 Feb 2014 07:19:04 +0000 (02:19 -0500)]
Implement the QR algorithm for computing eigenvalues.
Michael Orlitzky [Tue, 4 Feb 2014 07:18:09 +0000 (02:18 -0500)]
Add type synonyms for column/row matrices.
Add row' and column' functions that do what you'd expect only they return Mats instead of Vecs like their un-prime counterparts.
Michael Orlitzky [Mon, 3 Feb 2014 06:03:14 +0000 (01:03 -0500)]
Small cleanups in Linear.Vector.
Add 'diagonal' and 'trace' functions to Linear.Matrix.
Michael Orlitzky [Mon, 3 Feb 2014 02:10:01 +0000 (21:10 -0500)]
Clean up imports everywhere.
Michael Orlitzky [Mon, 3 Feb 2014 01:21:33 +0000 (20:21 -0500)]
Add tolerant versions of is_{upper,lower}_triangular.
Fix the QR factorization code.
Add tests for the QR behavior.
Michael Orlitzky [Sun, 2 Feb 2014 23:46:48 +0000 (18:46 -0500)]
Add a Normed instance for (Vec n a).
Michael Orlitzky [Sun, 2 Feb 2014 23:46:05 +0000 (18:46 -0500)]
Switch Linear.Vector to the numeric prelude and add the element_sum function to it.
Use Linear.Vector.element_sum in Linear.Matrix.
Michael Orlitzky [Sun, 2 Feb 2014 20:41:58 +0000 (15:41 -0500)]
Remove old comment from Normed.
Michael Orlitzky [Sun, 2 Feb 2014 17:34:42 +0000 (12:34 -0500)]
Add the Frobenius norm to Linear.Matrix.
Michael Orlitzky [Sun, 2 Feb 2014 17:12:09 +0000 (12:12 -0500)]
Add the Linear.QR module and update the ghci/cabal files.
Michael Orlitzky [Sun, 2 Feb 2014 16:33:03 +0000 (11:33 -0500)]
New function: Linear.Matrix.identity_matrix.
Michael Orlitzky [Sun, 2 Feb 2014 04:16:01 +0000 (23:16 -0500)]
Fix implementation of Linear.Matrix.construct.
Michael Orlitzky [Thu, 30 Jan 2014 21:02:38 +0000 (16:02 -0500)]
Add the Polynomials.Orthogonal module.
Michael Orlitzky [Thu, 30 Jan 2014 14:19:17 +0000 (09:19 -0500)]
Bump dependencies.
Michael Orlitzky [Thu, 15 Aug 2013 17:11:43 +0000 (13:11 -0400)]
Add the rayleigh_quotient function.
Michael Orlitzky [Sun, 21 Jul 2013 21:21:12 +0000 (17:21 -0400)]
Add Gauss-Seidel and SOR iterative methods.
Michael Orlitzky [Sun, 21 Jul 2013 21:19:35 +0000 (17:19 -0400)]
Rename 'diagonal' to 'diagonal_part'.
Add functions to pull out the (strict) lower/upper triangular parts of a matrix.
Michael Orlitzky [Sun, 21 Jul 2013 18:25:30 +0000 (14:25 -0400)]
Fix a bunch of hlint warnings.
Michael Orlitzky [Sun, 21 Jul 2013 14:48:47 +0000 (10:48 -0400)]
Update ghci/cabal files to load the new module.
Michael Orlitzky [Sun, 21 Jul 2013 14:48:23 +0000 (10:48 -0400)]
Add the Linear.Iteration module.
Michael Orlitzky [Sun, 21 Jul 2013 14:14:19 +0000 (10:14 -0400)]
Add another example for forward_substitute.
Michael Orlitzky [Sun, 21 Jul 2013 14:13:50 +0000 (10:13 -0400)]
Add 'diagonal' function.
Michael Orlitzky [Sun, 21 Jul 2013 14:13:24 +0000 (10:13 -0400)]
Constrain Normed instance to column vectors (m-by-1 matrices).
Michael Orlitzky [Mon, 1 Jul 2013 20:00:59 +0000 (16:00 -0400)]
Add the unit_roundoff function.
Michael Orlitzky [Sat, 8 Jun 2013 21:19:36 +0000 (17:19 -0400)]
Update numeric-prelude and fixed-vector.
Use the (!) function from the new fixed-vector.
Add the Fast/Simple trisection method and a test.
Michael Orlitzky [Sun, 24 Feb 2013 14:55:10 +0000 (09:55 -0500)]
Get determinants working.
Michael Orlitzky [Sun, 24 Feb 2013 07:41:03 +0000 (02:41 -0500)]
Replace the whole Matrix implementation with something a little better.
Add the Linear.System module.
Get all doctests working again.
Update .ghci/cabal files.
Michael Orlitzky [Sun, 24 Feb 2013 07:01:24 +0000 (02:01 -0500)]
Add Vec1, Vec4, Vec5 aliases.
Michael Orlitzky [Sun, 24 Feb 2013 00:05:26 +0000 (19:05 -0500)]
Add triangular functions, determinant, minor (all half-baked) to Linear.Matrix.
Michael Orlitzky [Sun, 24 Feb 2013 00:05:10 +0000 (19:05 -0500)]
Add the Linear.System module.
Michael Orlitzky [Sun, 24 Feb 2013 00:04:51 +0000 (19:04 -0500)]
Add the delete function to Vector.hs.
Michael Orlitzky [Thu, 21 Feb 2013 04:08:56 +0000 (23:08 -0500)]
Fix compiler warnings and doctests.
Michael Orlitzky [Thu, 21 Feb 2013 03:34:29 +0000 (22:34 -0500)]
Convert scalars to 1x1 vectors.
Michael Orlitzky [Thu, 21 Feb 2013 02:47:50 +0000 (21:47 -0500)]
A huge pile of crap upon Matrix/Vector.
Michael Orlitzky [Thu, 21 Feb 2013 02:47:15 +0000 (21:47 -0500)]
Update .ghci.
Michael Orlitzky [Thu, 21 Feb 2013 02:46:56 +0000 (21:46 -0500)]
Remove assumptions on the Normed class.
Michael Orlitzky [Thu, 21 Feb 2013 00:06:10 +0000 (19:06 -0500)]
Convert Integration/Simpson.hs and Integration/Trapezoid.hs to numeric-prelude.
Michael Orlitzky [Wed, 20 Feb 2013 23:56:13 +0000 (18:56 -0500)]
Use RealField/RealRing where possible instead of their constituents.
Convert src/ODE/IVP.hs to numeric-prelude.
Add RealRing/RealField instanced to BigFloat.
Michael Orlitzky [Wed, 20 Feb 2013 23:35:05 +0000 (18:35 -0500)]
Rename Aliases.hs to BigFloat.hs, now containing numeric-prelude instances for BigFloats.
Rewrite Normed.hs, Roots/Fast.hs, and Roots/Simple.hs for numeric-prelude.
Michael Orlitzky [Wed, 20 Feb 2013 19:31:41 +0000 (14:31 -0500)]
Make the partition function work with numeric-prelude.
Michael Orlitzky [Wed, 20 Feb 2013 19:31:33 +0000 (14:31 -0500)]
Add numeric-prelude to the deps.
Michael Orlitzky [Tue, 12 Feb 2013 17:16:39 +0000 (12:16 -0500)]
Move the Vector and Matrix modules under Linear.