Adding Prereqs, Instructors, Loc, and GEs

.gitignore

@@ -2,6 +2,7 @@
 __pycache__/
 *.py[cod]
 *$py.class
+.DS_Store
 
 # C extensions
 *.so

app.py

@@ -7,7 +7,10 @@
 from scipy import spatial
 import os 
 import json
-from recommend_GloVe.recommend_GloVe_average import recommend
+from functools import wraps
+import time
+from recommend_tfidf.recommend_tfidf import recommend
+# from recommend_GloVe.recommend_GloVe_average import recommend
 
 
 with open("./static/courses.json", "r") as courses_file:
@@ -16,23 +19,42 @@
 
 app = Flask(__name__)
 
+def timer(request):
+	@wraps(request)
+	def wrapper(*args, **kwargs):
+		start = time.perf_counter()
+		response = request(*args, **kwargs)
+		end = time.perf_counter()
+		run = end - start
+		print(f"Time to find a recommendation: {run} sec")
+		return response
+	return wrapper
+
 @app.route('/')
 def index():
-    return render_template('index.html') 
+    return render_template('index.html')
 
 @app.route('/rec',methods=['POST'])
+@timer
 def getvalue():
 	try:
-		coursename = request.form['search'].split(" ")[0]
-		lowlvl = 'lowlvl' in request.form
-		dept_filter = 'dept_filter' in request.form
-		df = recommend([coursename], blacklist_lowerlevel=lowlvl, blacklist_dept=dept_filter)
-		return render_template('result.html', tables = df, course = coursename)
+		user_input = request.form['search']
+		upperlvl = 'upperlvl' in request.form
+		cmc = 'cmc' in request.form
+		pomona = 'pomona' in request.form
+		hmc = 'hmc' in request.form
+		scripps = 'scripps' in request.form	
+		pitzer = 'pitzer' in request.form
+		other = 'other' in request.form
+		df = recommend(user_input, upperlvl=upperlvl, cmc=cmc, pomona=pomona, hmc=hmc, scripps=scripps, pitzer=pitzer, other=other)
+		return render_template('result.html', tables = df, course = user_input)
 	except Exception as e:
-		error = "Invalid Course ID. Please Try Again"
+		print(e)
+		error = "Invalid Description. Please Try Again"
 		return render_template('index.html', error = error) 
 
 @app.route('/search', methods=['POST'])
+@timer
 def search():
 	term = request.form['q']
 	print ('term: ', term)

course/courses.py

@@ -0,0 +1,93 @@
+import requests, os, csv
+from dotenv import load_dotenv
+import pandas as pd
+
+# Creates list of all course areas offered at the 5Cs
+
+courseareas = requests.get(f'http://jicsweb.pomona.edu/api/courseareas')
+coursedict = {}
+for codes in courseareas.json():
+	coursedict.update({codes['Code']: codes['Description']})
+
+
+# Generates url for all valid course areas of this semester and returns json
+
+def reqCourseInfo(code):
+	payload = {}
+	load_dotenv()
+	api_key = os.environ.get("API_KEY", None)
+	payload["api_key"]=api_key
+	r = requests.get("https://jicsweb.pomona.edu/api/Courses/2023;FA/" + code, params = payload)
+	try:
+		return r.json()
+	except:
+		print(f"No courses offered in course area {code} this semester")
+
+
+# Calling above function to get all valid course areas this current semester
+valid = {}
+for code in coursedict.keys():
+	if reqCourseInfo(code) is not None:
+		valid[code] = coursedict[code]
+
+
+# Writing data to CSV
+
+header = ['Course Area', 'CourseCode', 'Name', 'Course Description', 'Faculty', 'Campus', 'MeetTime', 'Weekdays', 'Prerequisites']
+
+with open('courses.csv', 'w', encoding='UTF8', newline='') as f:
+	writer = csv.writer(f)
+	writer.writerow(header)
+	for current in valid:
+		for course in reqCourseInfo(current):
+			try:
+				CourseArea = valid[current]
+				CourseCode = course['CourseCode']
+				Name = course['Name']
+				Description = course['Description']
+				Faculty = []
+				reqs = Description.find("Prerequisite:")
+				reqs1 = Description.find("Prerequisites:")
+				if reqs != -1:
+					Prerequisites = Description[reqs + len("Prerequisite:"):]
+				elif reqs1 != -1:
+					Prerequisites = Description[reqs1 + len("Prerequisites:"):]
+				else:
+					Prerequisites = 'None'
+				if len(course['Instructors']) == 1:
+					Faculty = course['Instructors'][0]['Name']
+					if Faculty == ', taff':
+						Faculty = 'Staff'
+				else:
+					for instructor in course['Instructors']:
+						if instructor['Name'] == ', taff':
+							Faculty.append('Staff')
+						else:
+							Faculty.append(instructor['Name'])
+				Campus = []
+				MeetTime = []
+				Weekdays = []
+				if len(course['Schedules']) == 1:
+					Campus = course['Schedules'][0]['Campus']
+					MeetTime = course['Schedules'][0]['MeetTime']
+					Weekdays = course['Schedules'][0]['Weekdays']
+				else:
+					for schedule in course['Schedules']:
+						Campus.append(schedule['Campus'])
+						MeetTime.append(schedule['MeetTime'])
+						Weekdays.append(schedule['Weekdays'])
+			except:
+				print("Insufficient information on course")
+			data = [CourseArea, CourseCode, Name, Description, Faculty, Campus, MeetTime, Weekdays, Prerequisites]
+			writer.writerow(data)
+
+df = pd.read_csv('courses.csv')
+duplicates = df.loc[df.duplicated(subset=['CourseCode'], keep=False), :]
+print("duplicates=", duplicates)
+for idx, row in duplicates.groupby('CourseCode')['Course Area']:
+    concat = ', '.join(row.values)
+    df.loc[df['CourseCode'] == idx, 'Course Area(s)'] = concat
+    df.drop_duplicates(subset=['CourseCode'], keep='first', inplace=True)
+
+df.drop('Course Area', axis=1, inplace=True)
+df.to_csv('courses.csv', index=False)

recommend_tfidf/generate_tfidf.py

@@ -0,0 +1,31 @@
+# this file only needs to be run once to generate the tfidf vectors and save them to a pickle file
+
+import pandas as pd
+import numpy as np
+from sklearn.feature_extraction.text import TfidfVectorizer
+import pickle
+import re
+
+data = pd.read_csv('course/courses.csv')
+data = data.dropna(subset=['CourseCode', 'Course Description'])
+data = data[~data['Course Description'].str.contains(r'^[^\w]*$', flags=re.IGNORECASE)]
+
+# Combine 'Description', 'CourseCode', 'Name', and 'Faculty' into a single entry
+data['combined'] = data.apply(lambda row: ' '.join([str(row['Course Description']), str(row['CourseCode']), str(row['Name']), str(row['Faculty']), str(row['Campus']), str(row['Prerequisites'])]), axis=1)
+course_descriptions = data["combined"].tolist()
+course_titles = data["CourseCode"].tolist()
+
+tfidfvectorizer = TfidfVectorizer(analyzer='word', stop_words= 'english', ngram_range=(2, 5))
+vectors = tfidfvectorizer.fit_transform(course_descriptions)
+
+print("Length of course_titles:", len(course_titles))
+print("Length of vectors:", len(vectors.toarray()))
+
+with open("recommend_tfidf/vectorizer.pkl", "wb") as f:
+    pickle.dump(tfidfvectorizer, f)
+
+# Include "Name" and "Description" in the course_data DataFrame
+course_data = pd.DataFrame({"course_title": course_titles, "vector": list(vectors.toarray()), "Name": data["Name"], "Description": data["Course Description"], "Instructor": data["Faculty"], "Location": data["Campus"], "Prerequisites": data["Prerequisites"], "GE(s) Satisfied": data["Course Area(s)"]})
+print("course_data =", course_data)
+
+course_data.to_pickle("recommend_tfidf/course_data.pkl")

recommend_tfidf/recommend_tfidf.py

@@ -0,0 +1,78 @@
+import pickle
+from scipy.spatial.distance import cosine
+import numpy as np
+import pandas as pd
+import re
+
+with open("recommend_tfidf/vectorizer.pkl", "rb") as f:
+    loaded_vectorizer = pickle.load(f)
+
+with open("recommend_tfidf/course_data.pkl", "rb") as f:
+    loaded_course_data = pd.read_pickle(f)
+
+loaded_course_vectors = np.array(loaded_course_data["vector"].tolist())
+
+def similarity(vec1, vec2):
+    '''similarity(vec1, vec2) -> cosine similarity of vectors'''
+    return 1 - cosine(vec1, vec2)
+
+def recommend(user_input, rec_num=10, upperlvl=False, cmc=True, pomona=True, hmc=True, scripps=True, pitzer=True, other=True):
+    query_vector = loaded_vectorizer.transform([user_input]).toarray().ravel()
+    similarities = []
+    for i in range(loaded_course_vectors.shape[0]):
+        similarities.append(similarity(query_vector, loaded_course_vectors[i]))
+
+    # get the indices of the most similar courses
+    indices = np.argsort(similarities)[::-1]
+
+    recommendations = []
+
+    # create the table with the most similar courses and their details
+    permited_colleges = []
+    default_colleges = ["CM", "PO", "HM", "SC", "PZ"]
+    if cmc:
+        permited_colleges.append("CM")
+    if pomona:
+        permited_colleges.append("PO")
+    if hmc:
+        permited_colleges.append("HM")
+    if scripps:
+        permited_colleges.append("SC")
+    if pitzer:
+        permited_colleges.append("PZ")
+    if other:
+        permited_colleges.append("JP")
+        permited_colleges.append("KS")
+        permited_colleges.append("JM")
+        permited_colleges.append("CH")
+        permited_colleges.append("PPO")
+        permited_colleges.append("AF")
+
+    num_added_classes = 0
+    for index in indices:
+        course_code = loaded_course_data.iloc[index]["course_title"]
+        course_college = re.search(r"(\w+)-", course_code).group(1)
+        if (course_college not in permited_colleges):
+            continue
+        course_lvl = re.search(r"\d+", course_code).group(0)
+        if (upperlvl and int(course_lvl) < 100):
+            continue
+        if(num_added_classes >= rec_num):
+            break
+        num_added_classes += 1
+        similarity_score = similarities[index]
+        course_details = {
+            "title": loaded_course_data.iloc[index]["Name"],
+            "description": loaded_course_data.iloc[index]["Description"],
+            "prerequisites": loaded_course_data.iloc[index]["Prerequisites"],
+            "instructor": loaded_course_data.iloc[index]["Instructor"],
+            "location": loaded_course_data.iloc[index]["Location"],
+	    "ges": loaded_course_data.iloc[index]["GE(s) Satisfied"]		
+        }
+        recommendations.append((course_code, similarity_score, course_details))
+
+    return recommendations
+
+#table = recommend("machine")
+#for row in table:
+#     print(row)

templates/head.html

@@ -16,7 +16,7 @@
             source : function(request, response) {
                 $.ajax({
                     type: "POST",
-                    url : "https://www.p-recs.com/search",
+                    url : "http://127.0.0.1:5000/search",
                     dataType : "json",
                     cache: false,
                     data : {
@@ -33,4 +33,4 @@
             minLength : 1
         });
     });
-</script>
+</script>

templates/index.html

@@ -14,4 +14,4 @@
 		</div>
 	</div>
 </body>
-</html>
+</html>

templates/result.html

@@ -13,7 +13,6 @@
 
         <div class="results">
             <h3>Recs for {{ course }}</h3>
-
             {% for table in tables %}
             <div class="course-rec">
                 <button type="button" class="collapsible button">
@@ -33,7 +32,31 @@ <h3>Recs for {{ course }}</h3>
                         Prerequisites
                     </p>
                     <p>{% if table[2].prerequisites %}{{ table[2].prerequisites }}{% else %}None{% endif %}</p>
-                </div>
+                    <p class="uppercase-title">
+                        Instructors
+                    </p>
+		    {% if table[2].instructor %}
+		    <p>{{ table[2].instructor }}</p>
+		    {% else %}
+    		    <p>None</p>
+		    {% endif %}
+                    <p class="uppercase-title">
+                        Location
+                    </p>
+                    {% if table[2].location %}
+                    <p>{{ table[2].location }}</p>
+                    {% else %}
+                    <p>None</p>
+                    {% endif %}
+                    <p class="uppercase-title">
+                        GE(s) Satisfied
+                    </p>
+                    {% if table[2].ges %}
+                    <p>{{ table[2].ges }}</p>
+                    {% else %}
+                    <p>None</p>
+                    {% endif %}
+
             </div>
             {% endfor %}
         </div>

templates/search.html

@@ -12,12 +12,22 @@ <h1><a class="left-col" href="/">p-recs</a></h1>
     <input type="text" name="search" id="searchBox" placeholder="Input course name or ID">
     <!-- submit = button!; value = what shows up on the button-->
     <div class="checkboxes">
-        <input type="checkbox" name="lowlvl" value="1"/> Remove lower level courses
+        <input type="checkbox" name="upperlvl" value="1"/> Remove lower level courses
         <br>
-        <input type="checkbox" name="dept_filter" value="1"/> Remove same department courses
+        <input type="checkbox" checked name="cmc" value="1"/> Include CMC courses
+        <br>
+        <input type="checkbox" checked name="pomona" value="1"/> Include Pomona courses
+        <br>
+        <input type="checkbox" checked name="hmc" value="1"/> Include Harvey Mudd courses
+        <br>
+        <input type="checkbox" checked name="pitzer" value="1"/> Include Pitzer courses
+        <br>
+        <input type="checkbox" checked name="scripps" value="1"/> Include Scripps courses
+        <br>
+        <input type="checkbox" checked name="other" value="1"/> Include Other courses
     </div>
     <input type="submit" value="Get recs">
 </form>
 {% if error %}
 <p class="error"><strong>Error:</strong> {{ error }}
-    {% endif %}
+    {% endif %}