Handle ranges per guidelines

.github/workflows/run_eval_tests.yml → .github/workflows/run_tests.yml

@@ -1,4 +1,4 @@
-name: Run Unit Tests for Evaluation scripts via Pytest
+name: Run Unit Tests via Pytest
 
 on:
   push:

.gitignore

@@ -20,3 +20,6 @@ results
 
 # ignore geopy cache (used for normalizing locations faster)
 geopy_cache.sqlite
+
+# pytest
+.pytest_*

Database/fix_nested_json.py

@@ -1,10 +1,8 @@
 import argparse
 import pathlib
 
-#from Database.scr.normalize_utils import Logging, NormalizeJsonOutput
-# this one works tested by Ni 20240719
 from scr.normalize_utils import Logging, NormalizeJsonOutput
-#from .scr.normalize_utils import Logging, NormalizeJsonOutput
+
 if __name__ == "__main__":
     parser = argparse.ArgumentParser()
     logger = Logging.get_logger("fix nested json sys output")

Database/output/nlp4climate_guidelines_v2/README.md

@@ -0,0 +1,7 @@
+### Annotation Guidelines v2 (changes in the normalization of numbers)
+
+The files in this directory are the same gpt output from the [nlp4climate](Database/output/nlp4climate) experiments but with the normalization rules for a revised version of the annotation guidelines (v2).
+
+The [jupyter notebook HTML export](guidelines_v1_versus_v2.html) shows a comparison with all items that underwent a change in the dev and test sets for these two categories: `Damage` and `Deaths`.
+
+The annotation guidelines [can be found here](https://onedrive.live.com/personal/78d0e12ab2e8ce00/_layouts/15/doc2.aspx?resid=78D0E12AB2E8CE00!sb951b21f6a3b4408808a037df599c45d&cid=78d0e12ab2e8ce00&migratedtospo=true&app=Word) (you may need to request pormission to access the file).