epsilon-compositionality

commit 446e8c0ef8dd8162fa76826893e43c4db52ffe42
parent a02677026c33d217b07506ec99bfea99f20c1d63
Author: Sondre Wold <sondrewold1@gmail.com>
Date:   Tue, 26 Aug 2025 09:55:27 +0200

fast error handling

Diffstat:
M
experiment.py
 | 
12
+++++++++---

1 file changed, 9 insertions(+), 3 deletions(-)
diff --git a/experiment.py b/experiment.py
@@ -247,8 +247,14 @@ def find_diff(a: str, b: str) -> list[str] | None:
     if a == b:  # Should not happen for SICK or NCIP, but you never know.
         return None
 
-    words_a = a.split()
-    words_b = b.split()
+    #print(f"A={a}\nB={b}\n##############")
+
+    try:
+        words_a = a.split()
+        words_b = b.split()
+    except:
+        raise Exception(f"Parse error. A={a}, B={b}")
+
 
     prefix_len = 0
     min_len = min(len(words_a), len(words_b))
@@ -400,7 +406,7 @@ def main() -> None:
     elif args.experiment_type == "ncip-non":
         # Here we measure the difference between two non-idiomatic phrases in context.
         sentence_A = df["synonym both"].values.tolist()
-        sentence_B = df["both synonym alt2_1"].values.tolist()
+        sentence_B = df["both synonym alt1_1"].values.tolist()
 
     elif args.experiment_type == "sick":
         lexA = df.loc[df["sentence_A_expRule"] == "S2_lex"]