From b9efdcab67a29f7fbacaaeb20b71b8ce730b3d26 Mon Sep 17 00:00:00 2001
From: Dennis Sweeney <36520290+sweeneyde@users.noreply.github.com>
Date: Fri, 27 Feb 2026 11:39:19 -0500
Subject: [PATCH] Correct description of shuffled_by_action

---
 README.md | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/README.md b/README.md
index 6656c78..b9f917c 100644
--- a/README.md
+++ b/README.md
@@ -382,8 +382,8 @@ More generally, we can shuffle the entries of the `Vector`
 Using some function from `range(len(v))` to `range(result_size)`,
 represented as a "action" `Vector`.
 If `result_size` is not provided, it defaults to `len(v)`.
-If the has duplicate entries `a[j1] == a[j2] == ...`,
-then `v.shuffled_by_action(a)[a[j1]]` will be their sum `a[j1] + a[j2] + ...`,
+If the action `a` has duplicate entries `a[j1] == a[j2] == ...`,
+then `v.shuffled_by_action(a)[a[j1]]` will be their sum `v[j1] + v[j2] + ...`,
 so each action `Vector` specifies a linear operation.
 
 ```pycon
@@ -448,4 +448,4 @@ Lattice(3, [[3, 0, 0], [0, 4, 4]], maxrank=2)
 The single optional argument of `L.decompose(keep_together)` should be an iterable
 of "component" iterables, and the integers in each component iterable are then ensured
 to be placed in the same component. The summands are allocated "tightly",
-with `maxrank` equal to their actual rank.
\ No newline at end of file
+with `maxrank` equal to their actual rank.