Christoph Ruegg
|
6754045a56
|
LA: diagonal matrix-vector Multiply LeftMultiply/TransposeThisAndMultiply
|
13 years ago |
|
Christoph Ruegg
|
aadb8a4bb5
|
LA: diagonal-dense Multiply TransoseAndMultiply TransposeThisAndMultiply
|
13 years ago |
|
Christoph Ruegg
|
44228a6d41
|
LA: Build.SameType of one or two example instances
|
13 years ago |
|
Christoph Ruegg
|
2e1dc29f96
|
LA: dense-diagonal TransposeThisAndMultiply
|
13 years ago |
|
Christoph Ruegg
|
19873cb304
|
LA: dense-diagonal TransposeAndMultiply
|
13 years ago |
|
Christoph Ruegg
|
06be8470b6
|
LA: dense-diagonal Add Subtract Multiply
|
13 years ago |
|
Christoph Ruegg
|
3259ce6ba2
|
LA: extend diagonal matrix builder
|
13 years ago |
|
Christoph Ruegg
|
8b92117356
|
LA: Matrix.ClearSubMatrix should nop-return instead of fail on empty/negative selection
|
13 years ago |
|
Christoph Ruegg
|
13ede7dc3a
|
Linear Regression: use more efficient LA routines (via tibel)
|
13 years ago |
|
Christoph Ruegg
|
2853fe4b9c
|
Fit: weighted MultiDim and weighted Polynomial
|
13 years ago |
|
Christoph Ruegg
|
233b4d4bcc
|
Distance: add Minkowski distance (generalized p-norm)
|
13 years ago |
|
Christoph Ruegg
|
df9948b48f
|
Distance: add Canberra distance
|
13 years ago |
|
Christoph Ruegg
|
7133c1448b
|
Correlation: optimize Spearman ranked correlation algorithm #171
|
13 years ago |
|
Christoph Ruegg
|
4f3a9a4f6d
|
Statistics: comment in Correlation.Person on robustness
|
13 years ago |
|
Christoph Ruegg
|
0d2a5b096c
|
GoodnessOfFit: docs, add R value
|
13 years ago |
|
Christoph Ruegg
|
08ca8301b8
|
Distance: add Pearson's distance
|
13 years ago |
|
Christoph Ruegg
|
3ba8773b30
|
GoodnessOfFit: benchmarking indicates pearson correlation is twice as fast
|
13 years ago |
Ethar Alali
|
6dda2ef725
|
Added license comments to file.
|
13 years ago |
|
Ethar Alali
|
8334454fad
|
Added RSquared class (and test - in TD like you mean it style) to the project.
|
13 years ago |
|
Christoph Ruegg
|
1b888c4d28
|
Tests: disable de-CH related number format tests because of Win 8.1 bug
|
13 years ago |
|
Christoph Ruegg
|
85db0a6884
|
Managed Provider: cache-optimized matrix L-infinity norm (8-10 times faster)
|
13 years ago |
|
Christoph Ruegg
|
2b41b9a969
|
Statistics: float-overloads for array Min/Max
|
13 years ago |
|
Christoph Ruegg
|
4e031e7a91
|
Statistics: drop redundant explicit null checks
|
13 years ago |
|
Christoph Ruegg
|
2a509edab3
|
Providers: Leverage MKL in Thin-QR also for single, complex and complex32 types.
|
13 years ago |
|
Christoph Ruegg
|
bff2f68396
|
LA: norml xml docs
|
13 years ago |
|
Christoph Ruegg
|
89c6824114
|
Build: roll back F# portable FSharp.Core hint path
|
13 years ago |
|
Christoph Ruegg
|
89c76125c6
|
Build: VS12 (2013) support, build config simplification
|
13 years ago |
|
Christoph Ruegg
|
798343ad73
|
Statistics: correlation matrix (pearson, spearman). Resolves #161.
|
13 years ago |
|
Christoph Ruegg
|
ee9d754172
|
Fit: more direct and slightly faster implementation; add MultiDim
|
13 years ago |
|
Christoph Ruegg
|
c7bd5c795a
|
MKL: Make bit-consistency optional and disabled by default
Disabled by default mainly for backwards compatibility
|
13 years ago |
|
Christoph Ruegg
|
48d03189be
|
Providers: drop unused matrix-norm work-array version, simplifications
|
13 years ago |
|
Christoph Ruegg
|
9e2b83551c
|
LA: Iterative solvers: make failure stop criterium generic/shared
|
13 years ago |
|
Christoph Ruegg
|
77e60dd62a
|
LA: Iterative solvers: make divergence stop criterium generic/shared
|
13 years ago |
|
Christoph Ruegg
|
db8a7a030c
|
LA: Iterative solvers: make residual stop criterium generic/shared
|
13 years ago |
|
Christoph Ruegg
|
be4583d133
|
Precision: rework comparisons, similar to equality before
|
13 years ago |
|
Christoph Ruegg
|
778b8151a5
|
Providers: double-valued matrix norms
|
13 years ago |
|
Christoph Ruegg
|
59b1005d7f
|
LA/Precision: Matrix norms double-valued (like vectors); generic AlmostEqual support
|
13 years ago |
|
Christoph Ruegg
|
afc4f7e167
|
Precision: rework/cleanup (major)
|
13 years ago |
Martin Posch
|
44f3541f42
|
Set Improved MKL Consistency
|
13 years ago |
|
Christoph Ruegg
|
8bc882bf88
|
Precision: cleanup
|
13 years ago |
|
Christoph Ruegg
|
8f38ca706a
|
Precision: fix inconsistent behavior
|
13 years ago |
|
Christoph Ruegg
|
728dc322ec
|
Docs: Turn F# Sample project into documentation (using FSharp.Literate/Formatting)
|
13 years ago |
tibel
|
3999c5ebc2
|
Fit: use generic classes and optimize Polynomial() matrix creation
|
13 years ago |
|
Christoph Ruegg
|
fadc4c87a0
|
Samples: extend F# distribution samples to Sample module, random walk
|
13 years ago |
|
Christoph Ruegg
|
6fb602cc0b
|
Samples: migrate F# random & distributions to FSharp.Formatting style
|
13 years ago |
|
Christoph Ruegg
|
9a5ecea3ac
|
Distributions: update F# module to core changes; easier to specify just seed
|
13 years ago |
|
Christoph Ruegg
|
6d7b0e3f2e
|
Random: use a less problematic seed by default
|
13 years ago |
|
Christoph Ruegg
|
ae053e168b
|
Control: fix portable build
|
13 years ago |
|
Christoph Ruegg
|
46bf65d87e
|
Control: simpler UseNativeMKL() and UseManaged() methods
|
13 years ago |
|
Christoph Ruegg
|
024570b9e6
|
Fix a negative-zero issue in complex numbers; make all portable tests pass
|
13 years ago |
