sequences
sequences copied to clipboard
haskell-perf
Benchmarks for sequence data structures: lists, vectors, etc.
Implements #27 through `criterion`. Since benchmarks already do `whnf/nf` - the benchmark simply runs `nf`.
[`tasty-bench`](https://hackage.haskell.org/package/tasty-bench) is a benchmarking library, that can be (more or less) used as a drop-in replacement for criterion, while having the advantage of being a lot faster, both to compile...
The recent activity by @Anton-Latukha has brought me to update the "acc" package and its benchmark suite, which is focused on monoidal construction and reduction using various data-structures. See https://github.com/nikita-volkov/acc#benchmark-results...
Since benchmarks probably would really follow the data type authors approach "make counterproductive things unrepresentable" (like not providing features where data types are completely unsuited for according features, for example...
Also an interesting question. That got raised in https://github.com/nikita-volkov/acc/issues/4#issuecomment-975769302 As some data types are lazy - current tests seem to not differentiate between types that are created to be normalized...
During `DList` use I had a difficulty to reason about: https://github.com/haskell/haskell-language-server/blob/2e61b7c0ba64cded22cf0fb80b089eaf328528e6/ghcide/src/Development/IDE/Core/Compile.hs#L475-L483 Because [`DList`](https://hackage.haskell.org/package/dlist-1.0/docs/Data-DList.html) docs state: > fromList: ... This function is implemented with ++. Repeated uses of fromList are just...
`traverse pure` probably suitable to be used for data structure spine traversal "in the data-type-recommended way". Maybe there even more idiomatic way to do it. Maybe `traverse (const $ Identity...
Metadata
Owner
Metadata
Benchmarks for sequence data structures: lists, vectors, etc.