Add shuffling

strmlt.py

@@ -118,9 +118,11 @@ if __name__ == "__main__":
             )
             y_pred = flatten(f["y_pred"][:-1] for q in questions for f in forecasts[q])
 
-        y_true, y_pred = np.array(y_true), np.array(y_pred)
+    y_true, y_pred = np.array(y_true), np.array(y_pred)
 
-        st.write(f"Which gives us {len(y_pred)} datapoints to work with.")
+    order = np.arange(len(y_true))
+    np.random.default_rng(0).shuffle(order)
+    y_true, y_pred = y_true[order], y_pred[order]
 
         # ---