.

atlas/views.py

@@ -1169,7 +1169,7 @@ def collection_mark_question(request, pk, case_number):
             return redirect(
                 "atlas:collection_mark_question",
                 case_number=int(request.POST.get("goto")),
-                **kwargs
+                **kwargs,
             )
     else:
         for answer in answers:
@@ -1195,13 +1195,6 @@ def collection_mark_question(request, pk, case_number):
     )
 
 
-@user_is_collection_author_or_atlas_editor
-def collection_scores_cid(request, pk):
-    collection = get_object_or_404(CaseCollection, pk=pk)
-
-    return render(request, "atlas/collection_take.html", {"collection": collection})
-
-
 @user_is_collection_author_or_atlas_editor
 def collection_candidates(request, pk):
     collection = get_object_or_404(CaseCollection, pk=pk)
@@ -1324,7 +1317,7 @@ def collection_case_view_take(request, pk, case_number, cid, passcode):
             return redirect(
                 "atlas:collection_case_view_take",
                 case_number=int(request.POST.get("goto")),
-                **kwargs
+                **kwargs,
             )
     else:
         form = CidReportAnswerForm(instance=answer)
@@ -1404,3 +1397,146 @@ def collection_case_view(request, pk, case_number):
             "answer": answer,
         },
     )
+
+
+@user_is_collection_author_or_atlas_editor
+def collection_scores_cid(request, pk):
+    collection = get_object_or_404(CaseCollection, pk=pk)
+
+    if not collection.exam_mode:
+        raise Http404("Collection not in exam mode")
+
+    user_answers_and_marks = defaultdict(list)
+    user_answers_marks = defaultdict(list)
+    user_answers = defaultdict(list)
+    user_names = {}
+    # cid_passcodes = {}
+
+    by_question = defaultdict(dict)
+    unmarked = set()
+
+    cases = collection.cases.all().order_by("casedetail__sort_order").prefetch_related()
+
+    case_details = CaseDetail.objects.filter(
+        case__in=cases, collection=collection
+    ).prefetch_related("case")
+
+    cid_user_answers = CidReportAnswer.objects.filter(question__in=case_details)
+
+    # questions = exam.exam_questions.all().prefetch_related("answers")
+
+    # cid_user_answers = CidReportAnswer.objects.select_related("question").filter(
+    #    question__in=questions, exam__id=pk
+    # )
+
+    cids = set()
+
+    # Loop through all candidates
+    for cid_user_answer in cid_user_answers:
+        # Convoluted (probably...)
+        cid = cid_user_answer.cid
+        # cid_passcodes[cid] = cid_user_answer.passcode
+        cids.add(cid)
+        s = cid_user_answer
+        user_names[cid] = cid
+
+        q = cid_user_answer.question
+
+        # if not s:
+        #    # skip if no answer
+        #    user_answers_marks[cid].append(0)
+        #    user_answers[cid].append("")
+        #    by_question[q].append(("", 0))
+        #    continue
+
+        ans = s.answer
+
+        answer_score = s.get_answer_score()
+        if answer_score == "unmarked":
+            index = cases.index(q)
+            unmarked.add(index)
+        user_answers[cid].append(ans)
+        user_answers_marks[cid].append(answer_score)
+        user_answers_and_marks[cid].append((ans, answer_score))
+
+        by_question[q][cid] = (ans, answer_score)
+
+    user_scores = {}
+    user_scores_normalised = {}
+    for user in user_answers_marks:
+
+        user_scores[user] = sum(
+            [i for i in user_answers_marks[user] if i != "unmarked"]
+        )
+
+        user_scores_normalised[user] = user_scores[user]
+
+    # ignore scores of 0 for stats
+    user_scores_list = [i for i in user_scores.values() if i > 0]
+
+    max_score = len(cases)
+
+    if len(user_scores_list) < 1:
+        mean = 0
+        median = 0
+        mode = 0
+        fig_html = ""
+    else:
+        mean = statistics.mean(user_scores_list)
+        median = statistics.median(user_scores_list)
+        try:
+            mode = statistics.mode(user_scores_list)
+        except statistics.StatisticsError:
+            mode = "No unique mode"
+
+        df = user_scores_list
+        fig = px.histogram(
+            df,
+            x=0,
+            title="{}: distribution of scores".format(collection),
+            labels={"0": "Score"},
+            height=400,
+            width=600,
+        )
+        fig_html = fig.to_html()
+
+        collection.stats_mean = mean
+        collection.stats_median = median
+        collection.stats_mode = mode
+
+        collection.stats_candidates = len(user_scores_list)
+        collection.stats_max_possible = max_score
+
+        collection.stats_min = min(user_scores_list)
+        collection.stats_max = max(user_scores_list)
+
+        collection.stats_graph = fig_html
+
+        collection.user_scores = user_scores
+
+        collection.save()
+
+    return render(
+        request,
+        f"atlas/collection_scores.html",
+        {
+            "cids": sorted(cids),
+            # "cid_passcodes": cid_passcodes,
+            "exam": collection,
+            "unmarked": unmarked,
+            "questions": cases,
+            "by_question": by_question,
+            "user_answers": dict(user_answers),
+            "user_answers_marks": dict(user_answers_marks),
+            "user_scores": user_scores,
+            "user_scores_normalised": user_scores_normalised,
+            "user_scores_list": user_scores_list,
+            "user_names": user_names,
+            "user_answers_and_marks": user_answers_and_marks,
+            "max_score": max_score,
+            "mean": mean,
+            "median": median,
+            "mode": mode,
+            "plot": fig_html,
+        },
+    )