ox-arrays - Nested, compositional struct-of-arrays orthotope arrays

	Commit message (Collapse)	Author	Age
*	Run stylish-haskell	Mikolaj Konarski	31 hours
\|
*	Nested.hs: Fix typo in doc comment	Tom Smeding	31 hours
\|
*	Nested.hs: import and export ordering	Tom Smeding	31 hours
\|
*	Update doc comments in Lemmas	Tom Smeding	31 hours
\|
*	Move shape conversion ops to Data.Array.Nested.Convert	Mikolaj Konarski	31 hours
\|
*	Move code around in Data.Array.Nested.Convert	Mikolaj Konarski	32 hours
\|
*	Merge both Lemmas modules	Mikolaj Konarski	41 hours
\|
*	Eliminate the dependency cycle of Lemmas modules	Mikolaj Konarski	42 hours
\|
*	Move modules Permutation and Types	Mikolaj Konarski	42 hours
\|
*	Generalize some shape conversion functions	Mikolaj Konarski	2 days
\|
*	Revert "Remove the KnownNat constraint from (:5509)"	Tom Smeding	2 days
\| \| \| \| \| \| \|	This reverts commit 8890526cac9e6c4d5583d00fce55f32ba613cf31. Removing a _provided_ KnownNat constraint if we cannot remove the actual Nat information from the data type is rather pointless.
*	Shape/index function rename	Tom Smeding	2 days
\|
*	default-show-instances flag	Tom Smeding	2 days
\|
*	Remove the KnownNat constraint from (:5509)	Mikolaj Konarski	2 days
\|
*	Explicit re-exports of Base members	Tom Smeding	3 days
\| \| \| \| \|	This is to ensure that haddock doesn't insert a module reference, but displays the members of the Base module inline.
*	Flip arguments of mcastToShaped for consistency	Tom Smeding	3 days
\|
*	Convert: Export list; simplify some functions with castCastable	Tom Smeding	3 days
\|
*	Move casts to DAN.Convert; split Ranked/Shaped types into .Base	Tom Smeding	3 days
\|
*	Make mcast available in Castable	Tom Smeding	3 days
\|
*	mcastSafe was unsound; it's now gone	Tom Smeding	3 days
\|
*	Simplify lemRankApp	Tom Smeding	3 days
\|
*	Move Data.Array.Arith to Data.Array.Strided.Orthotope	Tom Smeding	3 days
\|
*	Move Arith, XArray and Convert	Mikolaj Konarski	4 days
\|
*	Rename the three main public tensor API modules	Mikolaj Konarski	4 days
\|
*	Add .stylish-haskell.yaml that matches Tom's preferences	Mikolaj Konarski	4 days
\| \| \| \| \|	to be automatically applied at buffer save or periodically applied with find -not -path '/\.git/' -name "*.hs" -type f\| xargs stylish-haskell -i
*	Split and uniformly rename Shape modules	Mikolaj Konarski	4 days
\|
*	Appease GHC 9.10.2	Tom Smeding	5 days
\|
*	Revert some stylistic changes	Tom Smeding	5 days
\| \| \| \|	https://x.com/comerijn/status/1257804634833420292
*	Fix some of the warnings GHC 9.12.2 now produces	Mikolaj Konarski	5 days
\|
*	Apply some suggestions from hlint	Mikolaj Konarski	5 days
\|
*	Let stylish-haskell sort things	Mikolaj Konarski	5 days
\|
*	Add a couple of new zip operations that obviously make sense	Mikolaj Konarski	5 days
\|
*	Generalise list{x,s}Zip{,With} from Const to f	Tom Smeding	6 days
\|
*	Let's not add commented redundant needless errors	Tom Smeding	6 days
\|
*	Add zip and zipWith for sized lists	Mikolaj Konarski	7 days
\|
*	Fill in missing Head and HEAD	Mikolaj Konarski	7 days
\|
*	Cargo-cult NFData for all sized list types	Mikolaj Konarski	7 days
\|
*	Define fooLength and/or fooRank whenever not yet defined	Mikolaj Konarski	7 days
\|
*	Fix the Show instance for Shaped	Mikolaj Konarski	2025-04-26
\|
*	Refactor the clever replicate-aware Show instances	Tom Smeding	2025-04-26
\|
*	Do away with ShowVia{ToListLinear,Primitive}, were unused	Tom Smeding	2025-04-26
\|
*	Show concisely arrays replicated from a single element (scalar or not)	Mikolaj Konarski	2025-04-26
\|
*	Add mstrideTree and StrideTree	Mikolaj Konarski	2025-04-25
\|
*	Improve an error message	Tom Smeding	2025-03-25
\|
*	Compatibility liftVEltwise1 (TODO remove)	Tom Smeding	2025-03-20
\|
*	Some re-exports for compatibility	Tom Smeding	2025-03-20
\|
*	Separate arith routines into a library	Tom Smeding	2025-03-20
\| \| \| \|	The point is that this separate library does not depend on orthotope.
*	Arith statistics collection from C	Tom Smeding	2025-03-18
\|
*	arith: Don't FFI-import unused dotprod_*_strided ops	Tom Smeding	2025-03-18
\|
*	Optimise reductions and dotprod with more vectorisation	Tom Smeding	2025-03-14
\| \| \| \| \| \| \| \| \|	Turns out that if you don't supply -ffast-math, the C compiler will faithfully reproduce your linear reduction order, which is rather disastrous for parallelisation with vector units. This changes the summation order, so numerical results might differ slightly. To wit: the test suite needed adjustment.