adding deployment for surrogate

Author: JBris

app/conf/deployments_form/common.yaml

@@ -534,7 +534,7 @@ components:
           children: Neural density calibrator
           color: primary
           className: me-1
-      - id: call-surrogate-modelling-button
+      - id: call-surrogate-button
         label: Surrogate model
         help: Make a web request to the deployed surrogate model calibrator
         class_name: dash_bootstrap_components.Button 

app/pages/deployments_root_system.py

@@ -487,6 +487,53 @@ def call_snpe(n_clicks: int | list[int], statistics_inputs: list[float]) -> Call
     return dcc.send_file(outfile), True, f"Calling {task} calibrator"
 
 
+@callback(
+    Output(f"{PAGE_ID}-download-results", "data", allow_duplicate=True),
+    Output(f"{PAGE_ID}-load-toast", "is_open", allow_duplicate=True),
+    Output(f"{PAGE_ID}-load-toast", "children", allow_duplicate=True),
+    Input({"index": f"{PAGE_ID}-call-surrogate-button", "type": ALL}, "n_clicks"),
+    State({"type": f"{PAGE_ID}-parameters", "index": ALL}, "value"),
+    prevent_initial_call=True,
+)
+def call_surrogate(
+    n_clicks: int | list[int], statistics_inputs: list[float]
+) -> Callable:
+    """Call the surrogate model endpoint.
+
+    Args:
+        n_clicks (int | list[int]):
+            The number of form clicks.
+        statistics_inputs (list[float]):
+            The list of summary statistic values.
+
+    Returns:
+        Callable:
+            The form data.
+    """
+    if n_clicks is None or len(n_clicks) == 0:  # type: ignore
+        return no_update
+
+    if n_clicks[0] is None or n_clicks[0] == 0:  # type: ignore
+        return no_update
+
+    endpoint = os.environ.get(
+        "DEPLOYMENT_SURROGATE_INTERNAL_LINK", "http://surrogate:3000"
+    )
+
+    app = get_app()
+    statistics_form = app.settings[FORM_NAME]
+    summary_statistics = {}
+    for i, child in enumerate(statistics_form.components["parameters"]["children"]):
+        statistic_name = child["param"]
+        statistic_value = statistics_inputs[i]
+        summary_statistics[statistic_name] = statistic_value
+
+    json = {"data": [summary_statistics]}
+    task = "surrogate"
+    outfile = endpoint_predict(task, endpoint, json)
+    return dcc.send_file(outfile), True, f"Calling {task} calibrator"
+
+
 ######################################
 # Layout
 ######################################