Remove numpy, scipy from deps, disable chi test

setup.py

@@ -48,8 +48,6 @@
     install_requires=[
         "Wagtail>=4.1",
         "user-agents>=2.2,<2.3",
-        "numpy>=1.19.4,<1.25",
-        "scipy>=1.5.4,<1.10",
     ],
     extras_require={
         "testing": ["dj-database-url==0.5.0", "freezegun==1.2.1"],

wagtail_ab_testing/migrations/0013_alter_abtest_variant_revision.py

@@ -0,0 +1,20 @@
+# Generated by Django 4.1.13 on 2023-12-08 05:03
+
+from django.db import migrations, models
+import django.db.models.deletion
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('wagtailcore', '0083_workflowcontenttype'),
+        ('wagtail_ab_testing', '0001_squashed_0012_abtest_variant_revision'),
+    ]
+
+    operations = [
+        migrations.AlterField(
+            model_name='abtest',
+            name='variant_revision',
+            field=models.ForeignKey(on_delete=django.db.models.deletion.PROTECT, related_name='+', to='wagtailcore.revision'),
+        ),
+    ]

wagtail_ab_testing/models.py

@@ -2,8 +2,6 @@
 
 from datetime import datetime, timedelta, timezone as tz
 
-import scipy.stats
-import numpy as np
 from django.conf import settings
 from django.core.validators import MinValueValidator
 from django.db import connection, models, transaction
@@ -274,56 +272,7 @@ def log_conversion(self, version, *, time=None):
         AbTestHourlyLog._increment_stats(self, version, 0, 1, time=time)
 
     def check_for_winner(self):
-        """
-        Performs a Chi-Squared test to check if there is a clear winner.
-
-        Returns VERSION_CONTROL or VERSION_VARIANT if there is one. Otherwise, it returns None.
-
-        For more information on what the Chi-Squared test does, see:
-        https://www.evanmiller.org/ab-testing/chi-squared.html
-        https://towardsdatascience.com/a-b-testing-with-chi-squared-test-to-maximize-conversions-and-ctrs-6599271a2c31
-        """
-        # Fetch stats from database
-        stats = self.hourly_logs.aggregate(
-            control_participants=Sum('participants', filter=Q(version=self.VERSION_CONTROL)),
-            control_conversions=Sum('conversions', filter=Q(version=self.VERSION_CONTROL)),
-            variant_participants=Sum('participants', filter=Q(version=self.VERSION_VARIANT)),
-            variant_conversions=Sum('conversions', filter=Q(version=self.VERSION_VARIANT)),
-        )
-        control_participants = stats['control_participants'] or 0
-        control_conversions = stats['control_conversions'] or 0
-        variant_participants = stats['variant_participants'] or 0
-        variant_conversions = stats['variant_conversions'] or 0
-
-        if not control_conversions and not variant_conversions:
-            return
-
-        if control_conversions > control_participants or variant_conversions > variant_participants:
-            # Something's up. I'm sure it's already clear in the UI what's going on, so let's not crash
-            return
-
-        # Create a numpy array with values to pass in to Chi-Squared test
-        control_failures = control_participants - control_conversions
-        variant_failures = variant_participants - variant_conversions
-
-        if control_failures == 0 and variant_failures == 0:
-            # Prevent this error: "The internally computed table of expected frequencies has a zero element at (0, 1)."
-            return
-
-        T = np.array([[control_conversions, control_failures], [variant_conversions, variant_failures]])
-
-        # Perform Chi-Squared test
-        p = scipy.stats.chi2_contingency(T, correction=False)[1]
-
-        # Check if there is a clear winner
-        required_confidence_level = 0.95  # 95%
-        if 1 - p > required_confidence_level:
-            # There is a clear winner!
-            # Return the one with the highest success rate
-            if (control_conversions / control_participants) > (variant_conversions / variant_participants):
-                return self.VERSION_CONTROL
-            else:
-                return self.VERSION_VARIANT
+        return None
 
     def get_status_description(self):
         """