Add Whisper to requirements, update setup.py to read from requirements.txt

- Add openai-whisper to requirements.txt for speech decoding
- Refactor setup.py to read install_requires from requirements.txt
- Update clustering.py to return NaN (not 0.5) when insufficient data
- Remove debug print statements from clustering.py
- Remove benchmark_cluster.py (temporary file)
- Add CLAUDE.md to .gitignore

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Author: jeremymanning

.gitignore

@@ -12,3 +12,4 @@ dist/*
 venv/
 docs/_build
 docs/sg_execution_times.rst
+CLAUDE.md

benchmark_cluster.py

@@ -96,9 +96,8 @@ def _get_weight_exact(egg, feature, distdict, permute, n_perms):
     rec = list(egg.get_rec_items().values[0])
 
     if len(rec) <= 2:
-        warnings.warn('Not enough recalls to compute fingerprint, returning default'
-              'fingerprint.. (everything is .5)')
-        return 0.5
+        warnings.warn('Not enough recalls to compute fingerprint, returning NaN')
+        return np.nan
 
     distmat = get_distmat(egg, feature, distdict)
 
@@ -174,14 +173,11 @@ def _get_weight_best(egg, feature, distdict, permute, n_perms, distance):
 
     rec = list(egg.get_rec_items().values[0])
     if len(rec) <= 2:
-        warnings.warn('Not enough recalls to compute fingerprint, returning default'
-              'fingerprint.. (everything is .5)')
+        warnings.warn('Not enough recalls to compute fingerprint, returning NaN')
         return np.nan
 
     distmat = get_distmat(egg, feature, distdict)
     matchmat = get_match(egg, feature, distdict)
-    print(f"DEBUG: matchmat.shape={matchmat.shape}, len(rec)={len(rec)}")
-    print(f"DEBUG: distmat.shape={distmat.shape}")
 
     ranks = []
     for i in range(len(rec)-1):

quail/analysis/clustering.py

@@ -4,3 +4,4 @@ matplotlib>=3.5.0
 seaborn>=0.12.0
 pandas>=2.0.0
 joblib>=1.3.0
+openai-whisper

requirements.txt

@@ -1,5 +1,6 @@
 # -*- coding: utf-8 -*-
 
+import os
 from setuptools import setup, find_packages
 
 DESCRIPTION = 'A python toolbox for analyzing and plotting free recall data'
@@ -12,11 +13,15 @@
 - Clustering metrics (e.g. single-number summaries of how often participants transition from recalling a word to another related word, where "related" can be user-defined.)
 - Many nice plotting functions
 - Convenience functions for loading in data
-- Automatically parse speech data (audio files) using wrappers for the Google Cloud Speech to Text API
+- Automatically parse speech data (audio files) using OpenAI Whisper
 
 The intended user of this toolbox is a memory researcher who seeks an easy way to analyze and visualize data from free recall psychology experiments.
 """
 
+# Read requirements from requirements.txt
+here = os.path.abspath(os.path.dirname(__file__))
+with open(os.path.join(here, 'requirements.txt')) as f:
+    requirements = [line.strip() for line in f if line.strip() and not line.startswith('#')]
 
 EXTRAS_REQUIRE={
     'speech-decoding': ["pydub", "openai-whisper"],
@@ -35,13 +40,6 @@
     license='MIT',
     packages=find_packages(exclude=('tests', 'docs', 'paper')),
     include_package_data=True,
-    install_requires=[
-        'numpy>=2.0.0',
-        'scipy>=1.10.0',
-        'matplotlib>=3.5.0',
-        'seaborn>=0.12.0',
-        'pandas>=2.0.0',
-        'joblib>=1.3.0',
-    ],
+    install_requires=requirements,
     extras_require=EXTRAS_REQUIRE,
 )