updates

Dockerfile.hf

@@ -1 +0,0 @@
-# HF Spaces–compatible

app/api/routes.py

@@ -44,14 +44,15 @@ async def predict_file(background_tasks: BackgroundTasks, file: UploadFile = Fil
     # Correctly get numeric drift scores per column
     _, drift_dict = run_drift_check(df[predictor.features], reference_df[predictor.features], "v1")
 
-    # Ensure minimal values for chart and safe numeric handling
+    # Ensure numeric drift values safe for frontend Plotly chart
     drift_for_chart = []
     for col, score in drift_dict.items():
         try:
             score_value = float(score)
-            score_value = max(score_value, 0.01)
+            if not np.isfinite(score_value):
+                score_value = 0.0
         except Exception:
-            score_value = 0.01
+            score_value = 0.0
         drift_for_chart.append({"column": col, "score": score_value})
 
     # Schedule full drift in background as before

app/monitoring/drift.py

@@ -1,4 +1,4 @@
-# Evidently logic
+# app/monitoring/drift.py
 import os
 import pandas as pd
 from evidently.report import Report
@@ -23,7 +23,7 @@ def run_drift_check(current_data: pd.DataFrame, reference_data: pd.DataFrame, mo
     """
     Run Evidently DataDriftPreset on current vs reference data,
     save HTML report, and run governance checks.
-    Returns a tuple: (alerts, report_metrics)
+    Returns a tuple: (alerts, drift_scores)
     """
     os.makedirs(REPORT_DIR, exist_ok=True)
 
@@ -34,10 +34,22 @@ def run_drift_check(current_data: pd.DataFrame, reference_data: pd.DataFrame, mo
     # Extract numeric drift scores per column
     report_dict = report.as_dict() if hasattr(report, "as_dict") else {}
     drift_scores = {}
-    for metric in report_dict.get("metrics", []):
-        if metric["metric"] == "DataDriftMetric":
-            for col_name, col_data in metric["result"].get("dataset_drift", {}).items():
-                drift_scores[col_name] = col_data.get("drift_score", 0.0)
+
+    metrics_list = report_dict.get("metrics", [])
+
+    for metric in metrics_list:
+        result = metric.get("result", {})
+        # Check column-level drift
+        drift_by_columns = result.get("drift_by_columns", {})
+        if drift_by_columns:
+            for col, info in drift_by_columns.items():
+                score = info.get("drift_score", 0.0)
+                if score is None or not pd.notna(score):
+                    score = 0.0
+                drift_scores[col] = float(score)
+        # fallback: Dataset-level drift metric (PSI share)
+        elif metric.get("metric") == "DatasetDriftMetric":
+            drift_scores["dataset"] = float(result.get("share_of_drifted_columns", 0.0))
 
     # Run governance checks (keeps existing alerts)
     alerts = governance.check_metrics(report_dict, model_version=model_version)

app/templates/dashboard.html

@@ -21,46 +21,45 @@
     <div id="drift-chart"></div>
 
 <script>
-    async function fetchResults(csvFile) {
-        const formData = new FormData();
-        formData.append("file", csvFile);
+async function fetchResults(csvFile) {
+    const formData = new FormData();
+    formData.append("file", csvFile);
 
-        const response = await fetch("/predict", {
-            method: "POST",
-            body: formData
-        });
-
-        const data = await response.json();
+    const response = await fetch("/predict", { method: "POST", body: formData });
+    const data = await response.json();
 
-        /* ✅ FIX: API returns `results`, not `predictions` */
-        document.getElementById("predictions").innerHTML =
-            `<pre>${JSON.stringify(data.results, null, 2)}</pre>`;
+    document.getElementById("predictions").innerHTML =
+        `<pre>${JSON.stringify(data.results, null, 2)}</pre>`;
 
-        const driftContainer = document.getElementById("drift-chart");
-        driftContainer.innerHTML = "";
+    const driftContainer = document.getElementById("drift-chart");
+    driftContainer.innerHTML = "";
 
-        if (Array.isArray(data.drift)) {
-            const cols = data.drift.map(d => d.column);
-            const scores = data.drift.map(d => d.score);
+    if (Array.isArray(data.drift)) {
+        const cols = data.drift.map(d => d.column);
+        const scores = data.drift.map(d => {
+            let val = Number(d.score);
+            if (!Number.isFinite(val)) val = 0;
+            return val;
+        });
 
-            Plotly.newPlot(driftContainer, [{
-                x: cols,
-                y: scores,
-                type: "bar"
-            }]);
-        } else {
-            driftContainer.innerHTML =
-                "<p>Drift report scheduled. Open the Evidently HTML report.</p>";
-        }
+        Plotly.newPlot(driftContainer, [{
+            x: cols,
+            y: scores,
+            type: "bar"
+        }]);
+    } else {
+        driftContainer.innerHTML =
+            "<p>Drift report scheduled. Open the Evidently HTML report.</p>";
     }
+}
 
-    document.getElementById("upload-form").addEventListener("submit", async (e) => {
-        e.preventDefault();
-        const fileInput = e.target.file.files[0];
-        if (fileInput) {
-            await fetchResults(fileInput);
-        }
-    });
+document.getElementById("upload-form").addEventListener("submit", async (e) => {
+    e.preventDefault();
+    const fileInput = e.target.file.files[0];
+    if (fileInput) {
+        await fetchResults(fileInput);
+    }
+});
 </script>
 </body>
 </html>

open_drift.py

@@ -1,3 +0,0 @@
-import webbrowser
-report_path = r"C:\Users\Rayquaza\Desktop\IT\ML Inference Service with Drift Detection\reports\evidently\drift_report.html"
-webbrowser.open(f"file://{report_path}")

scripts/evaluate.py

@@ -1 +0,0 @@
-# offline evaluation

scripts/run_drift_check.py

@@ -1 +0,0 @@
-# batch drift job