Show if datasets are identical (#4029)

sitespeedio · Dec 14, 2023 · 28bb0ae · 28bb0ae
1 parent e58f9d9
commit 28bb0ae
Show file tree

Hide file tree

Showing 2 changed files with 10 additions and 10 deletions.
diff --git a/lib/plugins/compare/pug/index.pug b/lib/plugins/compare/pug/index.pug
@@ -74,8 +74,8 @@ table
             td
               a(href=createGraphLink(groupName, metricName)) 
                 b #{groupName + '.' + metricName}
-            if values.statisticalTestU === "N/A"
-              td N/A
+            if values.statisticalTestU === "N/A" || values.statisticalTestU === "Datasets are identical"
+              td #{values.statisticalTestU}
             else
               td #{h.decimals(values.statisticalTestU)}
             td #{h.decimals(values.baseline.mean)}
@@ -84,7 +84,7 @@ table
             td #{h.decimals(values.current.median)}
             td #{h.decimals(values.baseline.stdev)}
             td #{h.decimals(values.current.stdev)}
-            if values.statisticalTestU === "N/A"
+            if values.statisticalTestU === "N/A" || values.statisticalTestU === "Datasets are identical"
               td No Test Conducted
             else
               td #{values.statisticalTestU < 0.05 ? 'Yes - ' + cliffDeltaHelper(values.cliffsDelta) : 'No'}
@@ -96,7 +96,7 @@ each metricGroup, groupName in compare.metrics
   each values, metricName in metricGroup
     - var fullMetricName = groupName + '.' + metricName
     - var metricId = fullMetricName.replace(/\./g, '_')
-    h3 #{fullMetricName} #{values.statisticalTestU === "N/A"? '' : values.statisticalTestU < 0.05 ? '(significant change)' : ''}
+    h3 #{fullMetricName} #{values.statisticalTestU === "N/A" || values.statisticalTestU === "Datasets are identical" ? '' : values.statisticalTestU < 0.05 ? '(significant change)' : ''}
     .ct-chart(id=`chart-${metricId}`)
       .ct-legend
         span.ct-legend-item

diff --git a/lib/plugins/compare/statistical.py b/lib/plugins/compare/statistical.py
@@ -2,16 +2,17 @@
 import json
 from scipy.stats import wilcoxon, mannwhitneyu
 
-
 def has_variability(sample):
     """Check if the sample has more than one unique value."""
     return len(set(sample)) > 1
 
-
 def perform_test(test_type, baseline, current, **kwargs):
     """Perform the statistical test based on the test type."""
     if not has_variability(baseline) or not has_variability(current):
-        return None, "No variability"
+        if baseline == current:
+            return None, "Datasets are identical"
+        else:
+            return None, "No variability"
 
     if test_type == 'wilcoxon':
         return wilcoxon(current, baseline, **kwargs)
@@ -20,7 +21,6 @@ def perform_test(test_type, baseline, current, **kwargs):
     else:
         raise ValueError("Invalid test type. Choose 'wilcoxon' or 'mannwhitneyu'.")
 
-
 input_data = json.loads(sys.stdin.read())
 options = input_data['options']
 test_type = options.pop('test_type')
@@ -31,8 +31,8 @@ def perform_test(test_type, baseline, current, **kwargs):
     group_results = {}
     for metric_name, metric_data in metrics.items():
         stat, p = perform_test(test_type, metric_data['baseline'], metric_data['current'], **options)
-        if p == "No variability":
-            group_results[metric_name] = {'statistic': "N/A", 'p-value': "N/A"}
+        if p == "No variability" or p == "Datasets are identical":
+            group_results[metric_name] = {'statistic': "N/A", 'p-value': p}
         else:
             group_results[metric_name] = {'statistic': stat, 'p-value': p}
     final_results[group_name] = group_results