Note Data.Sequence performance improvements
authorDavid Feuer <David.Feuer@gmail.com>
Fri, 20 May 2016 00:19:28 +0000 (20:19 -0400)
committerDavid Feuer <David.Feuer@gmail.com>
Fri, 20 May 2016 00:19:28 +0000 (20:19 -0400)
changelog.md

index b13b131..f054005 100644 (file)
   * Add `foldTree` for `Data.Tree`.
 
   * Slightly optimize `replicateA` and `traverse` for `Data.Sequence`.
+  
+  * Substantially speed up `splitAt` and (consequently) `zipWith` for
+   `Data.Sequence` by building the result sequences eagerly. The improvements
+    are greatest for small sequences, but meaningful even for long ones.
+    Reimplement `take` and `drop` to avoid building and then discarding trees.
 
   * Speed up `adjust` for `Data.Map`.