-
-
Notifications
You must be signed in to change notification settings - Fork 40
SciSmalltalk Contents
For the use of SciSmalltalk in Squeak you should first load the compatibility package Math-SqueakPrerequisite.
These main packages exist in SciSmalltalk:
lets you do calculations with Floats with an arbitrary precision.
tests and dependencies can be found in ConfigurationOfArbitraryPrecisionFloat.
defines complex numbers and their functions and is completely self-explanatory.
it is tested in Math-Tests-Complex.
it is used by Math-Quaternion and an extension package that combines Math-Complex with Math-DHB-Numerical
is essentially the code (there are some minor changes) of the book Object-Oriented Implementation of Numerical Methods by Didier H. Besset and covers among other things interpolation, zero finding of functions, integration, series, statistics, linear algebra, optimization. you can have a look at the table of contents at the above link. the book documents and explains the code nicely. a freely downloadable version of this book without the java parts can be found at https://github.com/SquareBracketAssociates/NumericalMethods/releases. the original essentially unchanged code, workable in pharo and squeak (yet incompatible with scismalltalk), can be found at http://www.smalltalkhub.com/#!/~hernan/NumericalMethods.
some tests are in Math-Tests-DHB-Numerical, which mainly uses the code examples in the book.
it is used by several extension packages.
defines the hyperbolic functions, their inverses and sinc.
solves differential equations.
tests are in Math-Tests-ODE and in Math-Accuracy-ODE and (speed-tests) in Math-Benchmarks-ODE
can be used to repeatedly find nearest neighbours.
tests are in Math-Tests-KDTree and (speed-tests) in Math-Benchmarks-KDTree.
defines quaternions and their functions and is self-explanatory.
it needs Math-Complex to work correctly.
it is tested in Math-Tests-Quaternion.
it is used by an extension package that combines Math-Quaternion with Math-DHB-Numerical.
defines several pseudo-random-number generators and is self-explanatory.
tests are in Math-Tests-Random
adds some methods used mainly for testing, a QR factorization and some methods based on that factorization.
it needs Math-DHB-Numerical
tests are in Math-Tests-DHB-wk
by subclassing Accuracy one can easily write tests for the numerical accuracy of methods.
tests and an example are in this package
which can fit parameters of a function to some data and additionally can be used to find the minimum of a function.
it needs Math-DHB-Numerical and Math-AccuracyFramework
tests are in Math-Tests-FunctionFit which additionally needs Math-DHB-wk
implements some interval-arithmetic.
it needs Math-DHB-Numerical, Math-DHB-wk, Math-Number-Extensions, SMark, Roassal2
tests are in Math-Tests-RealInterval and speed-Tests in Math-Benchmarks-RealInterval.
lets you calculate first and second derivatives
it needs Math-DHB-Numerical
tests are in Math-Tests-AutomaticDifferentiation
lets you do a fast Fourier transformation and its inverse transformation
it needs Math-Complex
tests are in Math-Tests-FastFourierTransformation which additionally needs Math-DHB-Numerical
lets you calc quantiles and is self-explanatory and documented in its tests
tests are in the package
it is used by Math-KernelSmoothing
estimates a non-parametric density from experimental data
tests are in the package
it needs Math-Quantile and Math-DHB-Numerical
does a two-sided Kolmogorov-Smirnov-test. KolmogorovSmirnov1Sample checks whether sample data are from a population with a given distribution, KolmogorovSmirnov2Sample checks whether two sample data are from the same population. package is self-explanatory and heavily commented.
it needs Math-DHB-Numerical
tests are in Math-Tests-KolmogorovSmirnov
and then there are Math-ComplexExtensions and Math-QuaternionExtensions which make complex and quaternion available in in some DHB calculations. Math-DHB-NumericalExtensions are essentialy user-defined renamings of some methods.