Merge branch 'main' of github.com:COMSOC-Community/easychair-extra

easychair_extra/read.py

@@ -241,7 +241,7 @@ def read_submission(
                 if row["corresponding?"] == "yes":
                     corresponding_authors[sub_id].append(person_id)
         df["authors_id"] = df.apply(
-            lambda df_row: sub_to_authors.get(df_row["#"], []), axis=1
+            lambda df_row: tuple(sub_to_authors.get(df_row["#"], [])), axis=1
         )
         df["corresponding_id"] = df.apply(
             lambda df_row: corresponding_authors.get(df_row["#"], []), axis=1

examples/papers_with_same_authors.py

@@ -0,0 +1,41 @@
+import os
+
+import pandas as pd
+
+from easychair_extra.read import read_submission, read_committee
+
+
+def main():
+    current_dir = os.path.dirname(os.path.abspath(__file__))
+
+    # Read the submission file
+    submissions = read_submission(
+        os.path.join(current_dir, "..", "easychair_sample_files", "submission.csv"),
+        author_file_path=os.path.join(current_dir, "..", "easychair_sample_files", "author.csv"),
+    )
+
+    # Add a column with the number of authors
+    submissions["num_authors"] = submissions.apply(
+        lambda x: len(x["authors_id"]), axis=1
+    )
+    submissions.sort_values("num_authors", inplace=True, ascending=False)
+
+    duplicated_authors = submissions[submissions.duplicated(subset=["authors_id"])][
+        "authors_id"
+    ].unique()
+
+    print(f"These author_sets have one or more submission: {duplicated_authors}")
+    for authors in duplicated_authors:
+        print("=" * 20 + f"\nAuthor set: {authors}")
+        with pd.option_context(
+            "display.max_rows", None, "display.max_columns", None, "display.width", 500
+        ):
+            print(
+                submissions[submissions["authors_id"] == authors][
+                    ["#", "title", "authors_id", "authors"]
+                ].to_string(index=False)
+            )
+
+
+if __name__ == "__main__":
+    main()

examples/plot_num_pc_per_area.py

@@ -0,0 +1,161 @@
+import pathlib
+from collections import Counter, defaultdict
+
+import csv
+import os.path
+
+import pandas as pd
+import matplotlib.pyplot as plt
+import seaborn as sns
+
+from easychair_extra.read import read_committee, read_topics
+
+AREA_NICKNAMES = {
+    "Fairness, Ethics, and Trust": "FAIR",
+    "Computer Vision": "CV",
+    "Constraints and Satisfiability": "CONSAT",
+    "Data Mining": "DATA",
+    "Knowledge Representation and Reasoning": "KRR",
+    "Humans and AI": "HUM",
+    "Machine Learning": "ML",
+    "Multiagent Systems": "MAS",
+    "Natural Language Processing": "NLP",
+    "Planning and Search": "PLAN",
+    "Robotics": "ROBO",
+    "Uncertainty in AI": "UAI",
+    "Multidisciplinary Topics": "MULT",
+}
+
+AREA_NICKNAMES_SWAPPED = {
+    "FAIR": "Fairness, Ethics, and Trust",
+    "CV": "Computer Vision",
+    "CONSAT": "Constraints and Satisfiability",
+    "DATA": "Data Mining",
+    "KRR": "Knowledge Representation and Reasoning",
+    "HUM": "Humans and AI",
+    "ML": "Machine Learning",
+    "MAS": "Multiagent Systems",
+    "NLP": "Natural Language Processing",
+    "PLAN": "Planning and Search",
+    "ROBO": "Robotics",
+    "UAI": "Uncertainty in AI",
+    "MULT": "Multidisciplinary Topics",
+}
+
+ROLE_NICKNAMES = {
+    "associate chair": "AC",
+    "senior PC member": "SPC",
+    "chair": "PCC",
+    "PC member": "PC",
+}
+
+
+def compute_area_topic_to_weight(committee_df, topics_to_areas):
+    # We initialise some dicts that hold the information.
+    topics_to_pcs = defaultdict(lambda: defaultdict(list))  # topics => dict[role, list of PCs]
+    num_topics_per_pc = Counter()
+
+    def populate_pc_dicts(df_row):
+        pc_id = df_row["#"]
+        r = df_row["role"]
+        num_topics_per_pc[pc_id] += 1
+        for t in df_row["topics"]:
+            topics_to_pcs[t][r].append(pc_id)
+
+    # Populate the dicts initialised above
+    committee_df.apply(populate_pc_dicts, axis=1)
+
+    # We replace the PC members ids by their weight, for the topic dict
+    topics_to_weight = {}
+    for topic, role_dict in topics_to_pcs.items():
+        topics_to_weight[topic] = dict()
+        for role, pc_list in role_dict.items():
+            if role not in topics_to_weight[topic]:
+                topics_to_weight[topic][role] = sum(
+                    1 / num_topics_per_pc[pc] for pc in pc_list
+                )
+
+    # We sum up for the areas:
+    areas_to_weight = {}
+    for topic, role_dict in topics_to_weight.items():
+        area = topics_to_areas[topic]  # The area corresponding to the topic
+        if area not in areas_to_weight:
+            areas_to_weight[area] = dict()
+        for role, weight in role_dict.items():
+            if role not in areas_to_weight[area]:
+                areas_to_weight[area][role] = weight
+            else:
+                areas_to_weight[area][role] += weight
+
+    return areas_to_weight, topics_to_weight
+
+
+def plot_requirement_vs_practice(
+    area_weights,
+):
+    all_areas = sorted(AREA_NICKNAMES)
+    all_roles = ["associate chair", "senior PC member", "PC member"]
+
+    # Prepare the pandas frame
+    data = []
+    for area in all_areas:
+        for role in all_roles:
+            actual_value = area_weights[area].get(role, 0)
+            data.append(
+                {"area": area, "role": role, "type": "current", "value": actual_value}
+            )
+
+    df = pd.DataFrame(data)
+
+    plt.close("all")
+    g = sns.catplot(
+        data=df,
+        x="role",
+        y="value",
+        col="area",
+        col_wrap=3,
+        kind="bar",
+        sharey=False,
+        sharex=False,
+    )
+
+    # Add the values on top
+    for ax in g.axes.ravel():
+        for c in ax.containers:
+            labels = []
+            for v in c:
+                height = v.get_height()
+                if int(height) == height or height > 10:
+                    labels.append(str(round(height)))
+                else:
+                    labels.append(f"{height:.1f}")
+            ax.bar_label(c, labels=labels, label_type="edge")
+        ax.margins(y=0.2)
+
+    g.set_titles("{col_name}")
+    g.set_axis_labels("", "Count")
+
+    plt.show()
+
+
+def main():
+    current_dir = os.path.dirname(os.path.abspath(__file__))
+    root_dir = os.path.join(current_dir, "..", "easychair_sample_files")
+
+    # Read the committee file with the bids
+    committee_df = read_committee(
+        os.path.join(root_dir, "committee.csv"),
+        committee_topic_file_path=os.path.join(root_dir, "committee_topic.csv")
+    )
+
+    area_topic_mapping, topic_area_mapping = read_topics(os.path.join(root_dir, "topics.csv"))
+
+    # Compute the weight per area and per role
+    area_weight_map, topic_weight_map = compute_area_topic_to_weight(committee_df, topic_area_mapping)
+    plot_requirement_vs_practice(
+        area_weight_map,
+    )
+
+
+if __name__ == "__main__":
+    main()