Hipparchus Release Notes
Release History
| Version | Date | Description |
|---|---|---|
| 3.1 | 2024-04-05 | This is a maintenance release. It includes several bugfixes and improvements, and adds one important feature: constrained optimization! Other additions are Gauss-Markov order 1 autoregressive process generation and GCD/LCM to {Big}Fraction |
| 3.0 | 2023-10-08 | This is a major release. |
| 2.3 | 2022-10-05 | This is a maintenance release. |
| 2.2 | 2022-08-10 | This is a maintenance release. The main changes are: addition of Unscented Kalman Filter for non linear estimations and addition of Taylor map inversion. |
| 2.1 | 2022-04-24 | This is a maintenance release. The main changes are: addition of 3D version of complex function plotter, addition of ConvergenceCheckerOrMultiplexer and ConvergenceCheckerAndMultiplexer, addition of inverse Jacobi elliptic functions, optional modified weights to Akima interpolation, modified Gram-Schmidt basis orthonormalization process, generation of convergents streams in continued fractions, addition of equalsIee754 to points in all topologies to handle NaN coordinates. Various bugs were fixed. |
| 2.0 | 2021-08-07 | This is a major release. The main changes are the addition of a sequential Gauss-Newton optimizer, improved relative scheduling of step handlers and events handlers in ODE, addition of a finish callback in step handlers, replacement of RealFieldElement by CalculusFieldElement interface, many improvements in Complex to fully implement CalculusFieldElement with correct branch cuts so complex numbers can be used in many Hipparchus algorithms, addition of FieldComplex, addition of complex and field complex univariate integrals, addition of field univariate integrators, and faster simultaneous computation of sinhcos. Experimental implementation of elliptic functions and integrals for real, complex and general fields have been added. Beware the elliptic integrals for complex and field complex numbers in the incomplete case are considered experimental for now, they have known issues. Methods deprecated during the 1.X series have been removed. Numerous bugs were also fixed. |
| 1.8 | 2020-12-05 | This is a maintenance release. It fixes a few bugs, mainly a serious one for complex eigendecomposition when several eigenvalues are equal. It also add sum and angle differences in SinCos and FieldSinCos. |
| 1.7 | 2020-06-23 | This is a maintenance release. It fixes several bugs, including a nasty one in ODE events handling, improves a lot the Complex class so it implements the new CalculusFieldElement interface, and adds several new differentiation classes to handle simpler cases (univariate derivatives at orders 1 and 2 and gradients) with less overhead than the existing DerivativeStructure. |
| 1.6 | 2019-11-01 | This is a maintenance release. The main changes are the addition of a field version of Vector2D, addition of an Optimum.of() factory method and a visibility change of the AbstractEvaluation constructor. |
| 1.5 | 2019-05-06 | This is a maintenance release. The main changes since last version are the introduction of field versions of bilinear interpolation, univariate interpolators, polynomials and polynomial spline functions, and the introduction of reset events in ODE. |
| 1.4 | 2018-11-17 | This is a maintenance release. The main changes since last version are the introduction of bilinear interpolation for 2D grids, field versions of sinCos and a way to build three-dimensional field arrays, accessors for state transition matrix, Jacobian matrix H, innovation covariance matrix and Kalman gain in Kalman filters, support for complex ordinary differential equations (both primary and secondary equations), and various fixes in partitioning. |
| 1.3 | 2018-05-08 | This is a maintenance release. The main changes since last version are the introduction of complex eigen decomposition, introduction of a solver for continuous time algebraic Riccati equation, a new Kalman module featuring both a filter for linear processes and a filter for non-linear processes, capability for secondary equations to update the derivatives of the primary equation in ODE, and introduction of a matrix decomposer to configure decomposition threshold independently of the matrix. |
| 1.2 | 2017-10-25 | This is a maintenance release. |
| 1.1 | 2017-03-16 | This is a maintenance release. |
| 1.0 | 2016-06-23 | This is the first release of the library |
Release 3.1 – 2024-04-05
| Type | Changes | By |
|---|---|---|
| Add public, static access to integrators name. Fixes issues/269. | serrof | |
| Improved release guide. Fixes issues/277. Thanks to Bryan. | luc | |
| Added Gauss-Markov order 1 autoregressive process generation. Fixes issues/333. | luc | |
| Added equals and hashCode to OpenIntTo{Double|Field}HashMap. Fixes issues/289. | luc | |
| Allow variable arguments for building PolynomialFunction. Fixes issues/327. | luc | |
| Extracted class JacobiKey from PolynomialsUtils. Fixes issues/322. | maxime | |
| Allow retrieving correct Cardan/Euler angles even at singularity. Fixes issues/304. | luc | |
| Add possibility to use real coefficients for Butcher array based Field integrators. Fixes issues/300. | serrof | |
| Improved performance of Gradient. Fixes issues/302. | serrof | |
| Fixes regressions with Field in event detection. Fixes issues/290. | serrof | |
| Added constrained optimization. Fixes issues/296. Thanks to Francesco Rocca. | luc | |
| Added GCD and LCM to {Big}Fraction. Fixes issues/294. Thanks to Axel Kramer. | luc | |
| Replace uses of scalar multiplication on Field one. Fixes issues/288. | serrof | |
| Fix broken newInstance method for FieldTuple. Fixes issues/291. | serrof | |
| Rework interfaces for Derivative and FieldDerivative. Fixes issues/286. | serrof | |
| Add default implementations in CalculusFieldElement and inheritors. Fixes issues/281. | serrof | |
| Add square method to CalculusFieldElement. Fixes issues/280. | serrof |
Release 3.0 – 2023-10-08
| Type | Changes | By |
|---|---|---|
| Fixed internal error in RandomPercentile aggregation. Fixes issues/238. | luc | |
| Added Principal Component Analysis. Fixes issues/221. Thanks to Paul King. | luc | |
| Removed unused mapper in initializeStep. Fixes issues/271. Thanks to Romain Serra. | luc | |
| Max check interval in events detectors can now depend on current state. Fixes issues/270. | luc | |
| Fixed wrong intermediate frame in torque-free motion model. Fixes issues/230. | luc | |
| Added torque-free motion model for non-symmetrical rigid body in ODE tests, with analytical theoretical reference solution. Thanks to Lucas Girodet. | luc | |
| Separated {Field}ODEEventDetector and {Field}ODEEventHandler. Fixes issues/223. | luc | |
| Added reset state and stop integration capability at end of step for ODE integrators. Fixes issues/220. | luc | |
| Improve management of Cardan/Euler rotations close to singularity. Fixes issues/243. | luc | |
| Added Blendable interface and its implementation in RealMatrix and Vector interfaces. Changed Vector interface parameters. Added SmoothStepFactory which allow for quick creation of common and generic smoothstep function. Fixes issues/212. | vincent | |
| UnivariateDerivative now implements Comparable. Fixes issues/275. | serrof | |
| Complex now implements Comparable, using real part as primary sort order and imaginary part as secondary sort order (similar to what ComplexComparator already did). Fixes issues/274. | luc | |
| Fixed eigenvalues/eigenvectors sorting in EigenDecompositionSymmetric. Fixes issues/253. | luc | |
| Replaced EigenDecomposition by EigenDecompositionSymmetric and EigenDecompositionNonSymmetric. | luc | |
| Removed OrderedEigenDecomposition. | luc | |
| Added a protection against out of range BigFraction. Fixes issues/273. | luc | |
| Fixed compatibility with recent JDK. Fixes issues/272. | luc | |
| Made FieldPolynomialSplineFunction operational with Field not inheriting from Comparable. Fixes issues/257. Thanks to Romain Serra. | luc | |
| Fixed determinant computation in eigen decomposition with complex eigenvalues. Fixes issues/262. | luc | |
| Only remove input coefficients of FieldPolynomialFunction that satisfy isZero(). Fixes issues/259. Thanks to Romain Serra. | luc | |
| Improved performance for reciprocal, division and square root with DerivativeStructure and FieldDerivativeStructure. Fixes issues/256. Thanks to Romain Serra. | luc | |
| Generate zero vectors in OrderedComplexEigenDecomposition when the number of eigenvectors is smaller than the dimension. Fixes issues/249. | luc | |
| Allow customization of eigen value ordering in OrderedComplexEigenDecomposition. Fixes issues/248. | luc | |
| Fixed FastMath.expm1(x) for huge values of x. Fixes issues/228. | luc | |
| Renamed Decimal64{Field} into Binary64{Field}. Fixes issues/227. | luc | |
| Separated {Field}ODEEventDetector and {Field}ODEEventHandler. Fixes issues/223. | luc | |
| Fieldified Erf and Gamma classes and added FieldContinuedFraction. Fixes issues/218. | vcucchie |
Release 2.3 – 2022-10-05
| Type | Changes | By |
|---|---|---|
| Fixed wrong negation of point on the 2-sphere. Fixes issues/208. | luc | |
| Allow configuration of RandomPercentile in StreamingStatistics. Fixes issues/207. | luc | |
| Added new signature for SequentialGaussNewtonOptimizer#withAPrioriData allowing to define the Cholesky decomposition thresholds. Fixes issues/205. | bryan | |
| Changed HessenbergTransformer visibility to public. Fixes issues/204. | luc | |
| Changed SchurTransformer visibility to public. Fixes issues/204. | luc |
Release 2.2 – 2022-08-10
| Type | Changes | By |
|---|---|---|
| Added {Field}TaylorMap, supporting evaluation, composition and inversion. Fixes issues/190. | luc | |
| Added FieldMatrixDecomposer. Fixes issues/200. | luc | |
| Added rebase to {Field}DerivativeStructure, which enables changing the base variables a {Field}DerivativeStructure depends on. Fixes issues/191. | luc | |
| Added Unscented Kalman Filter for non linear estimations. Fixes issues/199. Thanks to Gaetan Pierre. | bryan | |
| Added CombinatoricsUtils.permutations. Fixes issues/198. | luc | |
| Added computation of Bell numbers. | luc | |
| Added RosenNumberPartitionIterator. Fixes issues/197. Thanks to Axel Kramer. | luc | |
| Added CombinatoricsUtils.partitions. Fixes issues/197. | luc | |
| Added FieldMultivariateFunction, CalculusFieldMultivariateFunction, FieldMultivariateVectorFunction, CalculusFieldMultivariateVectorFunction, FieldMultivariateMatrixFunction and CalculusFieldMultivariateMatrixFunction. Fixes issues/195. | luc | |
| Added a getPrimary method to FieldExpandableODE. Fixes issues/194. | luc | |
| Added integration/differentiation of {Field}DerivativeStructure. Fixes issues/192. Thanks to Romain Serra. | luc |
Release 2.1 – 2022-04-24
| Type | Changes | By |
|---|---|---|
| Fixed some wrong branch cuts in inverse Jacobi elliptic functions in the complex case. Fixes issues/189. | luc | |
| Added a 3D version of complex plots. | luc | |
| Added optional modified weights to avoid overshoots in Akima interpolation near constant slopes sub-samples. Fixes issues/172. | luc | |
| Added ConvergenceCheckerOrMultiplexer and ConvergenceCheckerAndMultiplexer. Fixes issues/136. | luc | |
| Avoid building an intermediate complex for computing double norm. Fixes issues/186. | luc | |
| Fixed overflows for incrementors that are allowed to reach Integer.MAX_VALUE. Fixes issues/175. | luc | |
| Added inverse Jacobi elliptic functions, for both double and field values. Fixes issues/188. | luc | |
| Fixed wrong set operation with ArcsSet when two BSP trees share the same angle as a cut hyperplane. Fixes issues/187. | luc | |
| Fixed non-bracketing issue when RESET_STATE slightly moves an event at the start of a step and another regular event happens in the first half of the same step. Fixes issues/184. | luc | |
| Added equalsIee754 to S1Point, S2Point, Vector1D, Vector2D and Vector3D to have an equality method consistent with IEEE754 concerning NaN coordinates, in addition to the regular equals method compliant with Java language design choices, including Double.equals (which is different from the double == operator). | luc | |
| Avoid too many calls to function in BracketingNthOrderBrentSolver when using AllowedSolution.ANY_SIDE. Fixes issues/183. | luc | |
| Added Modified Gram-Schmidt basis orthonormalization process in MatrixUtils for double-based and CalculusFieldElement-based list of vectors. Fixes issues/178. | luc | |
| Fixed issue in Laguerre solver related to different representations of signed zeros. Fixes issues/177. | luc | |
| Allow generation of continued fractions convergents as streams of BigFraction/Fraction to allow user to apply their own convergence criterion to select one convergent. Fixes issues/176. | luc | |
| Fixed inconsistency of eigenvalues and eigenvectors between OrderedComplexEigenDecomposition and base class ComplexEigenDecomposition. Fixes issues/173. | luc |
Release 2.0 – 2021-08-07
| Type | Changes | By |
|---|---|---|
| Added field versions of univariate integrators. Fixes issues/153. | luc | |
| Added numerical evaluation of complex elliptic integrals. | luc | |
| Added complex and field complex univariate integrals (including contour integrals). Fixes issues/149. | luc | |
| Allow retrieving event handler configuration from ODE integrators. Fixes issues/147. | luc | |
| Added a separate finish method in ODE step handlers to be called after last step. Fixes issues/146. | luc | |
| Fixed out-of-order scheduling of calls to step handlers and event handlers. Fixes issues/145. | luc | |
| Added ArrayFieldVector.toString(). Fixes issues/143. | luc | |
| Fixed array index error in field LU decomposition. Fixes issues/144. Thanks to Axel Kramer. | luc | |
| Added sequential Gauss-Newton optimizer. Thanks to Julie Bayard. | luc | |
| Added getRowDimension and getColumnDimension to DecompositionSolver and FieldDecompositionSolver. Fixes issues/135. | luc | |
| Fixed ulp() implementation for Dfp. Fixes issues/142. | luc | |
| Added a way to have a more accurate value of π for fields implementations that require it. Fixes issues/141. | luc | |
| Clarified arguments in the one-argument version of the two-arguments atan2 function in CalculusFieldElement. Fixes issues/140. | luc | |
| Renamed signum into sign in field classes, and improved implementation for complex numbers. Fixes issues/138. | luc | |
| Fixed consistent implementations between {Field}LUDecomposition classes. Fixes issues/131. | bryan | |
| Added Carlson elliptic integrals (R_F, R_J, R_G, R_D, and R_C) both for real, CalculusFieldElement, Complex, and FieldComplex. | luc | |
| Fixed branch cut on imaginary axis for complex atan. Fixes issues/129. | luc | |
| Changed return types for field elements methods (including Complex): norm() → double, abs() → field element. Fixes issues/124. | luc | |
| Added the four Jacobi theta functions θ₁(z|τ), θ₂(z|τ), θ₃(z|τ), and θ₄(z|τ), both for complex and field complex. | luc | |
| Added field-based complex numbers. Fixes issues/126. | luc | |
| Added Legendre elliptic integrals (K, K', E, D, F, Π), both complete and incomplete, both for real, CalculusFieldElement, Complex, and FieldComplex. | luc | |
| Added the twelve Jacobi elliptic functions sn, cn, dn, cs, ds, ns, dc, nc, sc, nd, sd, and cd, both for primitive double and fields. Fixes issues/119. | luc | |
| Added field implementations of ulp function. Fixes issues/123. | luc | |
| Added a way to compute both sinh and cosh at once faster than performing two calls. This also speeds up all derivatives implementations as well as complex computations. Fixes issues/122. | luc | |
| Removed methods and constants deprecated during the 1.x series. Fixes issues/121. | luc | |
| Replaced RealFieldElement by CalculusFieldElement interface and have Complex implement it so it can be used in many Hipparchus algorithms (like ODE). Also replaced RealField{Univariate|Bivariate|Vector|Matrix}Function by CalculusField{Univariate|Bivariate|Vector|Matrix}Function. Fixes issues/67. | luc | |
| Added event filtering for field ODE. Fixes issues/120. | luc | |
| Use primitive double for initial step and error factors in Field adaptive step size integrators. Fixes issues/118. | luc | |
| Added the Ryū algorithm to generate very fast the shortest decimal representation of a floating point number that maintains round-trip safety. | luc | |
| Fix interpolation of a pair of ODE states whose times are the same. Such a pair can appear when a RESET_STATE or RESET_DERIVATIVES event occurs at the integration stop time. Fixes issues/112. Thanks to andrewsgoetz. | andrewsgoetz | |
| Allow selecting customized threshold for tests in OrderedComplexEigenDecomposition. Fixes issues/111. | luc | |
| Added zero checker predicate for some linear algebra algorithms. Fixes issues/110. | luc | |
| Fixed exception in OrderedComplexEigenDecomposition when eigenvalues are equal. Fixes issues/109. | luc |
Release 1.8 – 2020-12-05
| Type | Changes | By |
|---|---|---|
| Added sum and difference of angles in SinCos and FieldSinCos. Fixes issues/106. | luc | |
| Removed spurious reduction of BigFraction in equality test. Fixes issues/105. | luc | |
| Fixed LU-decomposition issue with signed zero complex numbers on diagonal. Fixes issues/104. | luc | |
| Fixed complex eigendecomposition when several eigenvalues are equals. Fixes issues/99. | luc |
Release 1.7 – 2020-06-23
| Type | Changes | By |
|---|---|---|
| Added ExtendedUnivariateDifferentiableVectorFunction interface for differentiable vectorial function. Fixes issues/97. | bryan | |
| Added intermediate interface {Field}Derivative. The interface is implemented by both {Field}DerivativeStructure, {Field}Gradient, {Field}UnivariateDerivative1and {Field}UnivariateDerivative2 | luc | |
| Added new mathematical methods for FieldGradient, FieldUnivariateDerivative1 and FieldUnivariateDerivative2. Fixes issues/95. | bryan | |
| Added isMathematicalInteger to Complex. Fixes issues/92. | luc | |
| Added isMathematicalInteger to Precision for double and float numbers. | luc | |
| Fix infinite loop in event handling when a RESET_STATE or RESET_DERIVATIVES event causes two other event detectors to trigger simultaneously and discontinuously at the same time. Fixes issues/94. Thanks to Mikael. | evan | |
| Added UnivariateDerivative1, UnivariateDerivative2 and Gradient for computing derivatives in simple cases with less overhead than DerivativeStructure. Fixes issues/93. | luc | |
| Added map and mapToSelf to RealMatrix. Fixes issues/81. | luc | |
| Don't ignore exceptions, use try with resources. Fixes issues/71. | luc | |
| Fixed accuracy of linear combination for Dfp. Fixes issues/91. | luc | |
| Added Dfp conversion to either larger or smaller number of digits. | luc | |
| Fixed handling of special cases in Dfp.atan2(y, x). Fixes issues/88. | luc | |
| Prevent overflow of Dfp.hypot. Fixes issues/87. | luc | |
| Fixed Dfp.floor and Dfp.ceil for numbers close to zero. Fixes issues/86. | luc | |
| Fixed Decimal64.rootN for negative arguments and even roots. Fixes issues/82. | luc | |
| Provide both L_1 and L_infinity norms for matrices. Fixes issues/77. | luc | |
| Fixed MathArrays.linearCombination when signed zeros are involved. Fixes issues/76. | luc | |
| Add Line.fromDirection(...). Fixes issues/78. | evan | |
| Changed RotationOrder from final class to enum. Fixes issues/72. | bryan | |
| Fixed Complex.pow(Complex p) and Complex.pow(double p) for special cases (real arguments, integer power, 0). Fixes issues/74. | luc | |
| Added Complex.pow(int n) with integer power computed more accurately by successive squarings rather than complex logarithm and exponential. | luc | |
| Added signum to Fraction and BigFraction. Fixes issues/63. | luc | |
| Prevent random failure of micro-sphere test case. | luc |
Release 1.6 – 2019-11-01
| Type | Changes | By |
|---|---|---|
| Add Optimum.of(...). | ward | |
| Make AbstractEvaluation constructor public, fix javadoc. | ward | |
| Fixed missing safety check in multi-step integrators. Github: fixes #62. Fixes issues/62. | luc | |
| Added field version of Vector2D. | luc | |
| Prevent stack overflow error when attempting integration with too large number of points. Fixes issues/61. | luc | |
| Show full precision in exception message when root bracketing fails. Fixes issues/56. | ward | |
| When exception message formatting fails still show a helpful message and stack trace. Fixes issues/56. | ward |
Release 1.5 – 2019-05-06
| Type | Changes | By |
|---|---|---|
| Nullify corrected state after predict in abstract Kalman filter. Note that this affects only custom extensions of AbstractKalmanFilter because the implementations provided by Hipparchus (LinearKalmanFilter and ExtendedKalmanFilter) always call correction step after having called prediction step, so the nullified corrected state is reset properly at the end. This was initially reported as https://gitlab.orekit.org/orekit/orekit/issues/540. Thanks to Hao Peng. | luc | |
| Added FieldBivariateFunction and make both BilinearInterpolatingFunction and PiecewiseBicubicSplineInterpolatingFunction implement it. Github: fixes #55. Fixes issues/55. | luc | |
| Added FieldUnivariateInterpolator and make both LinearInterpolator, AkimaSplineInterpolator and SplineInterpolator implement it. | luc | |
| Added FieldPolynomialFunction and FieldPolynomialSplineFunction. | luc | |
| Avoid too many multi-step integrator restarts after event resetting derivatives. Github: fixes #54. Fixes issues/54. | luc | |
| Add Action.RESET_EVENTS to ask the integrator to recheck all event detectors without recomputing the step. Use when the g function of one event detector depends on events occurring in another event detector. | evan | |
| Fixed documentation for interpolation algorithms. Github: fixes #53. Fixes issues/53. | luc | |
| Fixed lines/polyhedrons intersection occurring on facets boundaries. This was ported from Apache Commons Geometry issue 38. Thanks to Matt Juntunen. | luc |
Release 1.4 – 2018-11-17
| Type | Changes | By |
|---|---|---|
| Added bilinear interpolation for 2D grids. Github: fixes #48. Fixes issues/48. | luc | |
| Added field version of sinCos. Github: fixes #47. Fixes issues/47. | luc | |
| Fixed 2D outline extraction from 3D polyhedrons sets with holes. GitHub: fixes #45. Fixes issues/45. | luc | |
| Added support for complex ordinary differential equations (both primary and secondary equations). Github: fixes #39. Fixes issues/39. | luc | |
| Fixed wrong construction of an Illinois solver that really built a Pegasus solver. Github: fixes #44. Fixes issues/44. | luc | |
| Renamed scale parameter to location parameter in the log-normal distribution, to avoid confusion. Github: fixes #4. Fixes issues/4. | luc | |
| Added accessors for state transition matrix, Jacobian matrix H, innovation covariance matrix and Kalman gain in Kalman filters. GitHub: fixes #43. Fixes issues/43. | luc | |
| Use topological information from BSP tree to connect more reliably edges and vertices when building the boundary loops of SphericalPolygonsSet. GitHub: fixes #42. Fixes issues/42. | luc | |
| Forbid use of unrealistic tolerances (below ulp(2π)) in spherical geometry. GitHub: fixes #42. Fixes issues/42. | luc | |
| Fixed a corner case in BSPTree split. GitHub: fixes #41. Fixes issues/41. | luc | |
| Bisection solvers failed to raise an exception when the search interval did not bracket a root. GitHub: fixes #40. Fixes issues/40. | luc | |
| Added a way to build three-dimensional field arrays. | luc |
Release 1.3 – 2018-05-08
| Type | Changes | By |
|---|---|---|
| Added in FastMath the methods Java 9 added in StrictMath. This ensures FastMath is always a drop-in replacement for Math/StrictMath. Github: fixes #36. Fixes issues/36. | luc | |
| Added FieldUnivariateVectorFunction and FieldUnivariateMatrixFunction interfaces, as well as RealFieldUnivariateVectorFunction and RealFieldUnivariateMatrixFunction Github: fixes #35. Fixes issues/35. | luc | |
| Added FieldUnivariateFunction interface and have polynomials implement it. Github: fixes #34. Fixes issues/34. | luc | |
| Added new utility methods to create real and field vectors from dimension. | luc | |
| Allow FastMath to operate on RealFieldElements. This simplifies a lot "fieldification" of algorithms previously implemented with primitive double numbers. | luc | |
| Added a transposeMultiply method to compute A^T.B without creating the intermediate transposed matrix. | luc | |
| Added a multiplyTransposed method to compute A.B^T without creating the intermediate transposed matrix. | luc | |
| Added linear and extended Kalman filters, with support for outliers rejection. Fixes issues/32. | luc | |
| Added MatrixDecomposer interface that encapsulates algorithms tolerances. Thanks to Evan Ward. | luc | |
| Added a way to compute both sin and cos at once faster than performing two calls. This also speeds up DerivativeStructure.sin and DerivativeStructure.cos. | luc | |
| Allow secondary equations to update derivatives of primary equation in ODE. GitHub: fixes #31. Fixes issues/31. | luc | |
| Fixed a problem with eigen decomposition of matrices with large coefficients. Github: fixes #30. Fixes issues/30. | luc | |
| Added a solver for Continuous time Algebraic Riccati Equation. Fixes issues/29. Thanks to Alessandro Romero. | luc | |
| Added complex eigen decomposition. Thanks to Alessandro Romero. | luc | |
| Corrected MannWhitneyTest to report the U statistic used in the test and improved accuracy of the reported p-values by 1) adding an implementation of exact computation for small sample sizes 2) applying continuity correction to the normal approximation when this is used 3) adjusting variance for the presence of ties in the data. The problem with the U statistic was reported against Apache Commons Math as MATH-1453. Fixes issues/38. | psteitz | |
| Corrected WilcoxonSignedRankTest to discard tied pairs and to correctly perform continuity correction when estimating p-values using the normal approximation. This was reported against Apache Commons Math as MATH-1233. Fixes issues/37. | psteitz |
Release 1.2 – 2017-10-25
| Type | Changes | By |
|---|---|---|
| Added new RealFieldElement implementations computing a tuple. This allow to perform the same computation on several values at once. The getReal() value correspond to the first component of the tuple. | luc | |
| Added a method to compute the orientation of a triplet of points in 2D. | luc | |
| Fixed an error in EmpiricalDistribution causing cumulative probabilities to be returned as NaN when some data bins are empty. This was reported against Apache Commons Math as MATH-1431. Github: fixes #25. Fixes issues/25. | psteitz | |
| Added bracketing helper methods in UnivariateSolverUtils for RealFieldUnivariateFunction. Github: fixes #24. Fixes issues/24. | luc | |
| Added a getBRep method to PolyhedronsSet to extract a Boundary REPresentation in regular cases (unbounded PolyhedronsSet or PolyhedronsSet with holes in their boundary are not supported). | luc | |
| Fixed NullPointerException when computing size of unbounded polygons sets. Fixes issues/23. | luc | |
| Improved performance of getRow and setRow for Array2DRowRealMatrix and Array2DRowFieldMatrix. Github: fixes #22. Fixes issues/22. Thanks to Christoph Dibak. | luc | |
| Fixed polygons creation from lists of vertices (both in Euclidean and spherical geometry) when separate edges share the same hyperplane (line or circle) but with different orientations. Fixes issues/21. | luc |
Release 1.1 – 2017-03-16
| Type | Changes | By |
|---|---|---|
| Core: Added support for sampling from enumerated distributions in RandomDataGenerator. GitHub: fixes issue #19. Fixes issues/19. | psteitz | |
| Core: Added support for exact integration in PolynomialFunction. GitHub: fixes issue #18. Fixes issues/18. | psteitz | |
| Added a factory to simplify building DerivativeStructure instances. | luc | |
| Core: Added proper implementations of equals and hashCode to Fields so they can be used as keys in maps. | luc | |
| Core: Added a derivatives method in Hermite interpolator, similar to the one from FieldHermiteInterpolator. | luc | |
| Core: Added field-based derivative structures. | luc | |
| Core: Added field-based QR decomposition. GitHub: fixes issue #17. Fixes issues/17. Thanks to Andrea Antolino. | luc | |
| Core: Performance improvement for Array2DRowRealMatrix.getSubMatrix() (and also Array2DRowFieldMatrix.getSubMatrix()). GitHub: fixes issue #12. Fixes issues/12. Thanks to Christoph Dibak. | ebourg | |
| Core: Improved parameter validation for probabilities provided to enumerated distribution constructors. | psteitz | |
| Core: Added methods to expose the probability mass function to EnumeratedRealDistribution and EnumeratedIntegerDistribution. | psteitz | |
| Stat: Fixed overly-aggressive sufficient data check in OLS/GLS regression classes. When there is no intercept in the model, the number of observations can equal the number of regressors. GitHub: fixes #13. Fixes issues/13. | psteitz | |
| Stat: Fixed bug in 2-sided binomial test leading to inflated p-values in some cases. This was reported against Commons Math as MATH-1381. | psteitz | |
| Stat: Add implementation of RANDOM algorithm for streaming percentile computation. | psteitz | |
| Stat: Add default "toString()" implementation to "AbstractStorelessUnivariateStatistic". | tn | |
| Stat: "StorelessUnivariateStatistic", "DescriptiveStatistics" and "StreamingStatistics" now extend from "DoubleConsumer", allowing to easily use them within streams. | tn | |
| Ode: Fixed Field ODE integrators handling of secondary equations. GitHub: fixes #7. Fixes issues/7. | luc | |
| Geometry: Fixed numerical noise issues in Welzl smallest enclosing ball computation. Thanks to jerdmanGH for reporting the issue. GitHub: fixes #20. Fixes issues/20. | luc | |
| Geometry: Added field based 3D lines. | luc | |
| Geometry: Added utility methods for field-based rotations and 3D vectors. | luc |
Release 1.0 – 2016-06-23
| Type | Changes | By |
|---|---|---|
| Creation of a migration module and an associated script to help switching from Apache Commons Math to Hipparchus. | luc | |
| Simplified exception hierarchy. | luc | |
| Modularized the library into several components (hipparchus-core, hipparchus-clustering, hipparchus-fft, hipparchus-fitting, hipparchus-genetics, hipparchus-geometry, hipparchus-geometry, hipparchus-ode, hipparchus-optim, hipparchus-samples, hipparchus-stat). | tn | |
| Forked from the Apache Commons Math library. | psteitz |