Start implementating genetic algorithm

Pipfile

@@ -29,6 +29,7 @@ py = "*"
 django = "*"
 networkx = "*"
 social-auth-app-django = "*"
+pandas = "*"
 
 [pipenv]
 allow_prereleases = true

gatorgrouper/utils/constants.py

@@ -4,6 +4,7 @@
 ALGORITHM_ROUND_ROBIN = "rrobin"
 ALGORITHM_RANDOM = "random"
 ALGORITHM_GRAPH = "graph"
+ALGORITHM_GENETIC = "genetic"
 DEFAULT_METHOD = ALGORITHM_RANDOM
 DEFAULT_GRPSIZE = 3
 DEFAULT_NUMGRP = 3
@@ -24,5 +25,5 @@
 # Define configuration variables
 
 DEFAULT_GROUP_SIZE = 2
-WORKBOOK = "GatorGrouper (Responses)"
-WORKBOOK_CSV = WORKBOOK.replace(" ", "_") + ".csv"
+WORKBOOK = "GatorGrouper_(Responses).csv"
+# WORKBOOK_CSV = WORKBOOK.replace(" ", "_") + ".csv"

gatorgrouper/utils/group_genetic.py

@@ -5,7 +5,9 @@
 import random
 import math
 import numpy as np
-from colors import bold
+from gatorgrouper.utils import constants
+from gatorgrouper.utils import workbook
+# from colors import bold
 
 
 class Student:
@@ -78,16 +80,17 @@ def __str__(self):
 
 def create():
     """Create the groups of student"""
-    students_to_group = workbook.STUDENTS[:]
+
+    students_to_group = ["a","b","c","d"]
     random.shuffle(students_to_group)
 
     grouping = list()
 
-    for _ in range(workbook.GROUPING_SIZE):
+    for _ in range(DEFAULT_GRPSIZE):
         grouping.append(list())
 
     for index, student in enumerate(students_to_group):
-        grouping[index % workbook.GROUPING_SIZE].append(student)
+        grouping[index % DEFAULT_GRPSIZE].append(student)
 
     if grouping < 1:
         print("CREATED TOO SMALL GROUPING")

gatorgrouper/utils/mutations.py

@@ -0,0 +1,33 @@
+import sys
+import random
+from typing import List
+from gatorgrouper.utils import group_genetic
+
+
+def swap(grouping):
+    # print("MUTATION")
+    group_count = len(grouping)
+    # print("TOTAL GROUPS: {}".format(group_count))
+    # print("BEFORE: {}".format(grouping))
+    first, second = random.sample(range(len(grouping)), 2)
+    first_index = random.randrange(len(grouping[first]))
+    second_index = random.randrange(len(grouping[second]))
+    # print("swapping student {} in group {} with student {} in group {}".format(first_index, first, second_index, second))
+    temp = grouping[second][second_index]
+    grouping[second][second_index] = grouping[first][first_index]
+    grouping[second][second_index] = temp
+    # grouping[first][first_index], grouping[second][second_index] = grouping[second][second_index], grouping[first][first_index]
+    # print("AFTER: {}".format(grouping))
+
+    return grouping
+
+
+def multi_swap(grouping):
+    num_swaps = random.randrange(1, 6)
+    for _ in range(num_swaps):
+        grouping = swap(grouping)
+    return grouping
+
+
+def get():
+    return [swap, multi_swap]

gatorgrouper/utils/parse_arguments.py

@@ -52,6 +52,7 @@ def parse_arguments(args):
             constants.ALGORITHM_GRAPH,
             constants.ALGORITHM_ROUND_ROBIN,
             constants.ALGORITHM_RANDOM,
+            constants.ALGORITHM_GENETIC,
         ],
         default=constants.DEFAULT_METHOD,
         required=False,

gatorgrouper/utils/workbook.py

@@ -0,0 +1,71 @@
+"""Integrate GatorGrouper with Google Sheets."""
+
+import csv
+import math
+import logging
+import gspread
+import pandas as pd
+from oauth2client.service_account import ServiceAccountCredentials
+
+from gatorgrouper.utils import group_genetic
+from gatorgrouper.utils import constants
+
+
+EMAIL_COL = None
+PREFERENCES_COL = None
+SKILLS_COLS = set()
+
+STUDENTS = None
+GROUPING_SIZE = None
+
+def get(group_size):
+    """Retrieve data from Google Sheets and write to a CSV file."""
+
+    global EMAIL_COL
+    global PREFERENCES_COL
+    global SKILLS_COLS
+
+    # formatted_records = list()
+    # for entry in records:
+    #     formatted_entry = list()
+    #     for index, (question, response) in enumerate(entry.items()):
+    #         if question == 'Email Address':
+    #             EMAIL_COL = index - 1  # subtracting one because timestamp column not collected
+    #             formatted_entry.append(response)
+    #         elif "prefer" in question:
+    #             PREFERENCES_COL = index - 1
+    #             formatted_entry.append(response)
+    #         elif "skill" in question:
+    #             SKILLS_COLS.add(index - 1)
+    #             formatted_entry.append(response)
+    #     formatted_records.append(formatted_entry)
+
+    global STUDENTS
+    global GROUPING_SIZE
+
+    # EMAIL_COL = 0
+    # PREFERENCES_COL = 1
+    # SKILLS_COLS = [2, 3, 4, 5, 6]
+
+    DATA = pd.read_csv(constants.WORKBOOK_CSV, header=None)
+
+    EMAILS = DATA.iloc[:, EMAIL_COL]
+
+    STUDENTS = list()
+    for current_row, email in enumerate(EMAILS):
+        skills = list()
+        for skill_col in SKILLS_COLS:
+            skills.append(DATA.iat[current_row, skill_col])
+        preferences_str = DATA.iat[current_row, PREFERENCES_COL]
+
+        if isinstance(preferences_str, float) and math.isnan(preferences_str):
+            preferences = []
+        else:
+            preferences = preferences_str.replace(" ", "").split(",")
+
+        STUDENTS.append(Student(email, skills, preferences))
+
+    # for student in STUDENTS:
+    #     print(str(student) + "\n")
+
+    GROUPING_SIZE = math.floor(len(STUDENTS) / group_size)

gatorgrouper_cli.py

@@ -10,6 +10,8 @@
 from gatorgrouper.utils import group_graph
 from gatorgrouper.utils import display
 from gatorgrouper.utils import constants
+from gatorgrouper.utils import group_genetic
+from gatorgrouper.utils import mutations
 
 
 if __name__ == "__main__":  # pragma: no cover
@@ -62,6 +64,8 @@
                 objective_weights=GG_ARGUMENTS.objective_weights,
                 objective_measures=GG_ARGUMENTS.objective_measures,
             )
+        elif GG_ARGUMENTS.method == constants.ALGORITHM_GENETIC:
+            GROUPED_STUDENT_IDENTIFIERS = group_genetic.evolve(60, 0.33, 0.16, 0.18, 0.66, mutations.get())
         else:
             GROUPED_STUDENT_IDENTIFIERS = group_creation.group_random_num_group(
                 SHUFFLED_STUDENT_IDENTIFIERS, GG_ARGUMENTS.num_group