1 files changed, 54 insertions, 0 deletions
diff --git a/cmds/train_test.go b/cmds/train_test.go
index 5a2ecee..2a3bbeb 100644
--- a/cmds/train_test.go
+++ b/cmds/train_test.go
@@ -221,6 +221,60 @@ func TestTrainModelConsumedByScore(t *testing.T) {
 	}
 }
 
+// Val concentration: when one capture batch dominates the validation window,
+// the report must say so (share/count/stamp) and stdout must flag it.
+func TestTrainReportsValStampConcentration(t *testing.T) {
+	dir := t.TempDir()
+	labelsPath := filepath.Join(dir, "labels.jsonl")
+	syntheticLabels(t, labelsPath, 50, 50)
+
+	// Restamp the newest 25 rows to one captured_at: val-frac 0.3 puts
+	// 30 rows in val, 25 of them from this batch.
+	rows, err := core.ReadLabels(labelsPath)
+	if err != nil {
+		t.Fatal(err)
+	}
+	stamp := rows[len(rows)-1].CapturedAt
+	for i := len(rows) - 25; i < len(rows); i++ {
+		rows[i].CapturedAt = stamp
+	}
+	if err := rewriteLabels(labelsPath, "", rows); err != nil {
+		t.Fatal(err)
+	}
+
+	cmd := &TrainCommand{}
+	if err := cmd.Init([]string{labelsPath, "--val-frac", "0.3",
+		"--model-out", filepath.Join(dir, "m.json"),
+		"--report-out", filepath.Join(dir, "r.json")}); err != nil {
+		t.Fatalf("Init: %v", err)
+	}
+	var out bytes.Buffer
+	if err := cmd.Run(nil, &out); err != nil {
+		t.Fatalf("Run: %v", err)
+	}
+
+	body, err := os.ReadFile(filepath.Join(dir, "r.json"))
+	if err != nil {
+		t.Fatal(err)
+	}
+	var report trainReport
+	if err := json.Unmarshal(body, &report); err != nil {
+		t.Fatal(err)
+	}
+	if want := 25.0 / 30.0; report.ValStampShare != want {
+		t.Errorf("val_max_stamp_share = %v, want %v", report.ValStampShare, want)
+	}
+	if report.ValStampCount != 25 {
+		t.Errorf("val_max_stamp_count = %d, want 25", report.ValStampCount)
+	}
+	if !report.ValStamp.Equal(stamp) {
+		t.Errorf("val_max_stamp = %v, want %v", report.ValStamp, stamp)
+	}
+	if !strings.Contains(out.String(), "one batch shaped this validation") {
+		t.Errorf("stdout should flag the concentrated validation:\n%s", out.String())
+	}
+}
+
 // Guard: too few validation positives must error loudly.
 func TestTrainGuardFewValPositives(t *testing.T) {
 	dir := t.TempDir()