From 48b3ee71a853a4c2a060d8c933f7aba053024975 Mon Sep 17 00:00:00 2001
From: Andrew Ghazi <6763470+andrewGhazi@users.noreply.github.com>
Date: Fri, 4 Oct 2024 13:02:31 -0400
Subject: [PATCH] wording

---
 episodes/large_data.Rmd | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/episodes/large_data.Rmd b/episodes/large_data.Rmd
index 4bc0ab1..d803b5a 100644
--- a/episodes/large_data.Rmd
+++ b/episodes/large_data.Rmd
@@ -359,11 +359,11 @@ accurate, though RSVD is much faster for file-backed matrices.
 
 :::: challenge
 
-The uncertainty from approximation error is sometimes psychologically
-objectionable. "Why can't my computer just give me the right answer?" One way to
-alleviate this feeling is to quantify the approximation error on a small test
-set like the sce we have here. Using the `ExactParam()` class, visualize the
-error in PC1 coordinates compared to the RSVD results.
+The uncertainty from approximation error is sometimes aggravating. "Why can't my
+computer just give me the right answer?" One way to alleviate this feeling is to
+quantify the approximation error on a small test set like the sce we have here.
+Using the `ExactParam()` class, visualize the error in PC1 coordinates compared
+to the RSVD results.
 
 ::: solution
 This code block calculates the exact PCA coordinates. Another thing to note: PC vectors are only identified up to a sign flip. We can see that the RSVD PC1 vector points in the