Platform yearly statistics

src/apps/competitions/statistics.py

@@ -1,23 +1,163 @@
+"""
+This script is created to compute two types of statistics:
+    1. Overall platform statistics for a specified year
+    2. Overall published competitions statistics
+
+Usage:
+    Bash into django console
+    ```
+    docker compose exec django ./manage.py shell_plus
+    ```
+
+    For overall platform statistics
+    ```
+    from competitions.statistics import create_codabench_statistics
+    create_codabench_statistics(year=2024)
+
+    # if year is not specified, current year is used by default
+    # a csv file named codabench_statistics_2024.csv is generated in statistics folder (for year=2024)
+    ```
+
+    For overall published competitions statistics
+    ```
+    from competitions.statistics import create_codabench_statistics_published_comps
+    create_codabench_statistics_published_comps()
+
+    # a csv file named codabench_statistics_published_comps.csv is generated in statistics folder
+    ```
+"""
+
 # --------------------------------------------------
 # Imports
 # --------------------------------------------------
 import os
-from competitions.models import Competition
+from datetime import datetime
+from competitions.models import Competition, Submission, CompetitionParticipant
+from profiles.models import User
 
 # --------------------------------------------------
 # Setting constants
 # --------------------------------------------------
 BASE_URL = "https://www.codabench.org/competitions/"
 STATISTICS_DIR = "/app/statistics/"
-CSV_FILE_NAME = "codabench_competition_statistics.csv"
-CSV_PATH = STATISTICS_DIR + CSV_FILE_NAME
 
 
-def create_codabench_statistics():
+def create_codabench_statistics(year=None):
+    """
+    This function prepares a CSV file with different statistics per month
+    """
+
+    # Set year to current year if None
+    if year is None:
+        year = datetime.now().year
+
+    # Create statistics directory if not already createad
+    if not os.path.exists(STATISTICS_DIR):
+        os.makedirs(STATISTICS_DIR)
+
+    rows_dict = {}
+
+    # Initialize sets for tracking total of users, participants and submissions
+    total_users = set()
+    total_participants = set()
+    total_submissions = set()
+
+    # Loop over months
+    for month in range(1, 13):
+
+        # count total competitions
+        tota_competitions_count = Competition.objects.filter(created_when__year=year, created_when__month=month).count()
+        rows_dict.setdefault("total_competitions", []).append(tota_competitions_count)
+
+        # count public competitions
+        public_competitions_count = Competition.objects.filter(created_when__year=year, created_when__month=month, published=True).count()
+        rows_dict.setdefault("public_competitions", []).append(public_competitions_count)
+
+        # count private competitions
+        private_competitions_count = Competition.objects.filter(created_when__year=year, created_when__month=month, published=False).count()
+        rows_dict.setdefault("private_competitions", []).append(private_competitions_count)
+
+        # Count new users
+        new_users_count = User.objects.filter(date_joined__year=year, date_joined__month=month).count()
+        rows_dict.setdefault("new_users", []).append(new_users_count)
+
+        # Count total users (including the current month)
+        new_user_ids = set(User.objects.filter(date_joined__year=year, date_joined__month=month).values_list('id', flat=True))
+        total_users.update(new_user_ids)
+        rows_dict.setdefault("total_users", []).append(len(total_users))
+
+        # Count new participants
+        new_participants_count = CompetitionParticipant.objects.filter(competition__created_when__year=year, competition__created_when__month=month).count()
+        rows_dict.setdefault("new_participants", []).append(new_participants_count)
+
+        # Count total participants (including the current month)
+        new_participants_ids = set(CompetitionParticipant.objects.filter(competition__created_when__year=year, competition__created_when__month=month).values_list('id', flat=True))
+        total_participants.update(new_participants_ids)
+        rows_dict.setdefault("total_participants", []).append(len(total_participants))
+
+        # Count new submissions
+        new_submissions_count = Submission.objects.filter(created_when__year=year, created_when__month=month).count()
+        rows_dict.setdefault("new_submissions", []).append(new_submissions_count)
+
+        # Submissions per day = total submissions/30
+        submissions_per_day = 0
+        if new_submissions_count > 0:
+            submissions_per_day = int(new_submissions_count / 30)
+        rows_dict.setdefault("submissions_per_day", []).append(submissions_per_day)
+
+        # Count successful submissions (i.e., those that are finished)
+        successful_submissions = Submission.objects.filter(created_when__year=year, created_when__month=month, status=Submission.FINISHED).count()
+        rows_dict.setdefault("finished_submissions", []).append(successful_submissions)
+
+        # Count failed submissions (i.e., those that are failed)
+        failed_submissions = Submission.objects.filter(created_when__year=year, created_when__month=month, status=Submission.FAILED).count()
+        rows_dict.setdefault("failed_submissions", []).append(failed_submissions)
+
+        # Count total submissions (including the current month)
+        new_submissions_ids = set(Submission.objects.filter(created_when__year=year, created_when__month=month).values_list('id', flat=True))
+        total_submissions.update(new_submissions_ids)
+        rows_dict.setdefault("total_submissions", []).append(len(total_submissions))
+
+    # Set CSV file and path
+    CSV_FILE_NAME = f"codabench_statistics_{year}.csv"
+    CSV_PATH = STATISTICS_DIR + CSV_FILE_NAME
+
+    # Define month abbreviations
+    month_abbr = ["Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug", "Sep", "Oct", "Nov", "Dec"]
+
+    # Open the CSV file in write mode
+    with open(CSV_PATH, 'w') as output_file:
+        # Write the header row only once if the file is empty
+        if output_file.tell() == 0:
+            header = f"{year}; " + "; ".join(month_abbr) + "; Total \n"
+            output_file.write(header)
+
+        # Loop over each metric in the rows_dict and write the corresponding row
+        for metric, values in rows_dict.items():
+
+            # for total_users, total_participants, and total_submissions
+            # total is the last value
+            # for others total is the sum
+            if metric in ["total_users", "total_participants", "total_submissions"]:
+                total = values[-1]
+            else:
+                # Calculate the total for the metric (sum of all monthly counts)
+                total = sum(values)
+
+            # Create a row with the metric name followed by the values for each month
+            row = f"{metric}; " + "; ".join(map(str, values)) + f"; {total} \n"
+            output_file.write(row)
+
+
+def create_codabench_statistics_published_comps():
     """
     This function prepares a CSV file with all published competitions
     """
 
+    # Set CSV file and path
+    CSV_FILE_NAME = "codabench_statistics_published_comps.csv"
+    CSV_PATH = STATISTICS_DIR + CSV_FILE_NAME
+
     # Create statistics directory if not already createad
     if not os.path.exists(STATISTICS_DIR):
         os.makedirs(STATISTICS_DIR)